Files
modeling-app/rust/kcl-lib/e2e/executor/inputs/server-rack-heavy.kcl

1562 lines
55 KiB
Plaintext
Raw Normal View History

//Zoo server rack model with fully detailed vertical rails
const bendRad = 0.05
const thickness = 0.1046
const railHeight = 45 // u
const originStart = [4, 1, 5.114]
const depth = 48
const serverDepth = 31
const width = 21.53
// simple caster model at each corner
fn caster = (originStart) => {
const plane001c = {
origin: [
-(3.543 - 2.756) / 2 + originStart[0],
-(3.543 - 2.756) / 2 + originStart[1],
4.114
],
x_axis: [1.0, 0.0, 0.0],
y_axis: [0.0, 1.0, 0.0],
z_axis: [0.0, 0.0, 1.0]
}
const sketch001c = startSketchOn(plane001c)
|> startProfileAt([0, 0], %)
|> xLine(length = 3.543)
|> yLine(length = 3.543)
|> xLine(length = -3.543)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> hole(circle(center = [
(3.543 - 2.756) / 2,
(3.543 - 2.756) / 2
], radius = 8.8 / 2 / 25.4), %)
|> hole(circle(center = [
(3.543 - 2.756) / 2 + 2.756,
(3.543 - 2.756) / 2
], radius = 8.8 / 2 / 25.4 ), %)
|> hole(circle(center = [
(3.543 - 2.756) / 2,
(3.543 - 2.756) / 2 + 2.756
], radius = 8.8 / 2 / 25.4 ), %)
|> hole(circle(center = [
(3.543 - 2.756) / 2 + 2.756,
(3.543 - 2.756) / 2 + 2.756
], radius = 8.8 / 2 / 25.4 ), %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> extrude(length = -.25)
const sketch002c = startSketchOn(sketch001c, face = 'START')
|> startProfileAt([-originStart[0], 2.2 + originStart[1]], %)
2025-04-18 17:40:44 -05:00
|> arc(
angle_start = 30,
angle_end = 330,
radius = 3.2 / 2
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
|> extrude(length = 3.1)
const plane002c = {
origin: [
0 + originStart[0],
0.8 + originStart[1],
0
],
x_axis: [1, 0, 0],
y_axis: [0, 0, 1],
z_axis: [1, 0, 0]
}
const sketch003c = startSketchOn(plane002c)
|> circle(center = [0, 1.2], radius = 2.48 / 2)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const examplec = extrude(sketch003c, length = -1 - (3 / 16))
return examplec
}
// generate 4 casters
caster([
(3.543 - 2.756) / 2,
(3.543 - 2.756) / 2 + 2.75
])
caster([
-(3.543 - 2.756) / 2 + width - 2.756,
(3.543 - 2.756) / 2 + .4
])
caster([
(3.543 - 2.756) / 2,
serverDepth + 5 - ((3.543 - 2.756) / 2) - 2.2
])
caster([
-(3.543 - 2.756) / 2 + width - 2.756,
depth - ((3.543 - 2.756) / 2) - 2.756 - 1.7
])
// left wall
const plane001 = {
origin: [
originStart[0] - 4,
originStart[1] + 2,
4.114
],
x_axis: [0, 1, 0],
y_axis: [0, 0, 1],
z_axis: [1, 0, 0]
}
const sketch001l = startSketchOn(plane001)
|> startProfileAt([0, 0], %)
|> xLine(length = serverDepth + .8)
|> angledLine(angle = -45, endAbsoluteY = 1)
|> xLine(length = -serverDepth + 2 - .8, tag = $seg01)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude001l = extrude(sketch001l, length = 1)
const sketch002l = startSketchOn(plane001)
|> startProfileAt([serverDepth + .8, 0], %)
|> yLine(length = railHeight * 1.75 + 2)
|> angledLine(angle = -135, endAbsoluteX = serverDepth - 1 + .8)
|> yLine(length = -railHeight * 1.75)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude002l = extrude(sketch002l, length = 1)
const sketch003l = startSketchOn(plane001)
|> startProfileAt([
serverDepth + .8,
railHeight * 1.75 + 2
], %)
|> xLine(length = -serverDepth - .8, tag = $seg02)
|> angledLine(angle = -45, endAbsoluteY = railHeight * 1.75 - 1 + 2)
|> xLine(length = serverDepth - 2 + .8)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude003l = extrude(sketch003l, length = 1)
const sketch004l = startSketchOn(plane001)
|> startProfileAt([0, 0], %)
|> yLine(length = railHeight * 1.75 + 2)
|> angledLine(angle = 135, endAbsoluteY = railHeight * 1.75 + 2 - 1)
|> yLine(length = -railHeight * 1.75)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude004l = extrude(sketch004l, length = 1)
const sketch005l = startSketchOn(plane001)
|> startProfileAt([serverDepth - 1.25, 1], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [-serverDepth + 2.25, railHeight * 1.75], tag = $lineToIntersect4)
|> xLine(length = 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [serverDepth - 2.25, -railHeight * 1.75], tag = $lineToIntersect5)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude005l = extrude(sketch005l, length = 1)
const sketch006l = startSketchOn(plane001)
|> startProfileAt([1, 1], %)
|> angledLineThatIntersects(
angle = 70,
intersectTag = lineToIntersect4,
offset = 0,
)
|> angledLine(angle = -70, length = 1.414 )
|> angledLine(angle = 70 + 180, endAbsoluteY = 2 - 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude006l = extrude(sketch006l, length = 1)
const sketch007l = startSketchOn(plane001)
|> startProfileAt([
serverDepth - 1.2,
railHeight * 1.75 + 1
], %)
|> angledLineThatIntersects(
angle = 70,
intersectTag = lineToIntersect5,
offset = 0,
)
|> angledLine(angle = -70, length = 1.414 )
|> angledLine(angle = 70 + 180, endAbsoluteY = railHeight * 1.75 + 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude007l = extrude(sketch007l, length = 1)
// right wall
const plane002 = {
origin: [
originStart[0] - 4 + width - 1,
originStart[1] - 1,
4.114
],
x_axis: [0, 1, 0],
y_axis: [0, 0, 1],
z_axis: [1, 0, 0]
}
const sketch001w = startSketchOn(plane002)
|> startProfileAt([0, 0], %)
|> xLine(length = depth)
|> angledLine(angle = -45, endAbsoluteY = 1)
|> xLine(length = -depth + 2, tag = $seg01w)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude001w = extrude(sketch001w, length = 1)
const sketch002w = startSketchOn(plane002)
|> startProfileAt([depth, 0], %)
|> yLine(length = railHeight * 1.75 + 2)
|> angledLine(angle = -135, endAbsoluteX = depth - 1)
|> yLine(length = -railHeight * 1.75)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude002w = extrude(sketch002w, length = 1)
const sketch003w = startSketchOn(plane002)
|> startProfileAt([depth, railHeight * 1.75 + 2], %)
|> xLine(length = -depth, tag = $seg02w)
|> angledLine(angle = -45, endAbsoluteY = railHeight * 1.75 - 1 + 2)
|> xLine(length = depth - 2)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude003w = extrude(sketch003w, length = 1)
const sketch004w = startSketchOn(plane002)
|> startProfileAt([0, 0], %)
|> yLine(length = railHeight * 1.75 + 2)
|> angledLine(angle = 135, endAbsoluteY = railHeight * 1.75 + 2 - 1)
|> yLine(length = -railHeight * 1.75)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude004w = extrude(sketch004w, length = 1)
const sketch005w = startSketchOn(plane002)
|> startProfileAt([1, 40.6 + 1.75 / 2], %)
|> angledLine(angle = -23, length = 35.5 )
|> angledLine(angle = -23 + 90 + 45, length = 1.413 )
|> angledLine(angle = -23, endAbsoluteX = 1, tag = $lineToIntersect)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude005w = extrude(sketch005w, length = 1)
const sketch006w = startSketchOn(plane002)
|> startProfileAt([
1 + 35.5 * cos(23 * pi() / 180),
40.6 - (35.5 * sin(23 * pi() / 180)) + 1.75 / 2
], %)
|> angledLine(angle = -23 + 90, endAbsoluteX = depth - 1)
|> yLine(length = 2.56)
|> angledLineThatIntersects(
angle = -23 + 90 + 180,
intersectTag = lineToIntersect,
offset = 0,
tag = $lineToIntersect2,
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude006w = extrude(sketch006w, length = 1)
const sketch007w = startSketchOn(plane002)
|> startProfileAt([depth - 1, 60.65 + 1.75 / 2], %)
|> angledLine(angle = -23 + 180, length = 34.93 , tag = $lineToIntersect3)
|> angledLine(angle = 23 - 90, length = 1.414 )
|> angledLineThatIntersects(
angle = -23 + 180,
intersectTag = lineToIntersect2,
offset = 0,
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude007w = extrude(sketch007w, length = 1)
const sketch008w = startSketchOn(plane002)
|> startProfileAt([1, 41.7 + 1.75 / 2], %)
|> angledLineThatIntersects(
angle = -23 + 90,
intersectTag = lineToIntersect3,
offset = 0,
)
|> angledLine(angle = -23 - 45, length = 1.414 )
|> angledLineThatIntersects(
angle = -23 - 90,
intersectTag = lineToIntersect,
offset = 0,
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude008w = extrude(sketch008w, length = 1)
const sketch009w = startSketchOn(plane002)
|> startProfileAt([31.2, 33.3 + 1.75 / 2], %)
|> angledLine(angle = -23 - 45, length = 1.414 )
|> angledLine(angle = 90 - 23, length = 28 )
|> angledLine(angle = -23 + 45, length = -1.414 )
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude009w = extrude(sketch009w, length = 1)
const sketch010w = startSketchOn(plane002)
|> startProfileAt([31.2, 33.3 + 1.75 / 2], %)
|> angledLine(angle = -23 - 45, length = 1.414 )
|> angledLine(angle = 180 - 23, length = 28 )
|> angledLine(angle = -23 + 45, length = 1.414 )
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude010w = extrude(sketch010w, length = 1)
const sketch011w = startSketchOn(plane002)
|> startProfileAt([
31.2 - ((28 - 2) * cos(23 * pi() / 180)),
33.3 + (28 - 2) * sin(23 * pi() / 180) + 1.75 / 2
], %)
|> angledLine(angle = 90 - 23, length = 28 - 2 )
|> angledLine(angle = -23 - 45, length = -1.414 )
|> angledLine(angle = 90 - 23 + 180, length = 28 )
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude011w = extrude(sketch011w, length = 1)
const sketch012w = startSketchOn(plane002)
|> startProfileAt([
31.2 + (28 - 2) * sin(23 * pi() / 180),
33.3 + (28 - 2) * cos(23 * pi() / 180) + 1.75 / 2
], %)
|> angledLine(angle = 180 - 23, length = 28 - 2 )
|> angledLine(angle = -23 - 45, length = -1.414 )
|> angledLine(angle = -23, length = 28 )
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude012w = extrude(sketch012w, length = 1)
const sketch013w = startSketchOn(plane002)
|> startProfileAt([
1 + 4 * cos(23 * pi() / 180),
40.6 - (4 * sin(23 * pi() / 180)) + 1.75 / 2
], %)
|> angledLine(angle = -23, length = 1 )
|> angledLine(angle = -23 + 90, endAbsoluteX = 1)
|> yLine(length = 2.56)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude013w = extrude(sketch013w, length = 1)
const sketch014w = startSketchOn(plane002)
|> startProfileAt([
1 + 12 * cos(23 * pi() / 180),
40.6 - (12 * sin(23 * pi() / 180)) + 1.75 / 2
], %)
|> angledLine(angle = -23 - 90, length = 36 / 2 )
|> angledLine(angle = -23, length = 1 )
|> angledLine(angle = -23 - 90, length = -36 / 2 )
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude014w = extrude(sketch014w, length = 1)
const sketch015w = startSketchOn(plane002)
|> startProfileAt([
1 + (36 - 5.55 - 8) * cos(23 * pi() / 180),
40.6 - ((36 - 5.55 - 8) * sin(23 * pi() / 180)) + 1.75 / 2
], %)
|> angledLine(angle = -23 - 90, length = 36 / 2 )
|> angledLine(angle = -23, length = 1 )
|> angledLine(angle = -23 - 90, length = -36 / 2 )
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude015w = extrude(sketch015w, length = 1)
const sketch016w = startSketchOn(plane002)
|> startProfileAt([
1 + (36 - 5.5) * cos(23 * pi() / 180),
40.6 - ((36 - 5.5) * sin(23 * pi() / 180)) + 1.75 / 2
], %)
|> angledLine(angle = -23 - 90, length = 36 / 2 )
|> angledLine(angle = -23, length = 1 )
|> angledLine(angle = -23 - 90, length = -36 / 2 )
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude016w = extrude(sketch016w, length = 1)
const sketch017w = startSketchOn(plane002)
|> startProfileAt([
depth - 1 - (3.7 * cos(23 * pi() / 180)),
60.65 + 3.7 * sin(23 * pi() / 180) + 1.75 / 2
], %)
2025-04-18 17:40:44 -05:00
|> arc(angleStart = -23, angleEnd = 180 - 23, radius = 7 / 2 + 2)
|> angledLine(angle = -23 + 180, length = -1 )
2025-04-18 17:40:44 -05:00
|> arc(angleStart = 180 - 23, angleEnd = -23, radius = 7 / 2 + 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude017w = extrude(sketch017w, length = 1)
const sketch018w = startSketchOn(plane002)
|> startProfileAt([
depth - 1 - (19.3 * cos(23 * pi() / 180)),
60.65 + 19.3 * sin(23 * pi() / 180) + 1.75 / 2
], %)
2025-04-18 17:40:44 -05:00
|> arc(angleStart = -23, angleEnd = 180 - 23, radius = 7 / 2 + 2)
|> angledLine(angle = -23 + 180, length = -1 )
2025-04-18 17:40:44 -05:00
|> arc(angleStart = 180 - 23, angleEnd = -23, radius = 7 / 2 + 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude018w = extrude(sketch018w, length = 1)
const sketch019w = startSketchOn(plane002)
|> startProfileAt([1, 27.8 + 1.75 / 2], %)
|> angledLine(angle = -23, length = 7 )
|> angledLine(angle = -23 + 90, length = -1 )
|> angledLine(angle = -23, endAbsoluteX = 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude019w = extrude(sketch019w, length = 1)
const sketch020w = startSketchOn(plane002)
|> startProfileAt([
1 + (36 - 5.53 - 12) * cos(23 * pi() / 180),
27.8 + 1.75 / 2 - ((36 - 5.53 - 12) * sin(23 * pi() / 180))
], %)
|> angledLine(angle = -23, length = 7 )
|> angledLine(angle = -23 + 90, length = -1 )
|> angledLine(angle = -23 + 180, length = 7 )
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude020w = extrude(sketch020w, length = 1)
const sketch021w = startSketchOn(plane002)
|> startProfileAt([1, 21.9], %)
|> angledLine(angle = -23, endAbsoluteX = depth - 1)
|> yLine(length = -1.1)
|> angledLine(angle = -23, endAbsoluteX = 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude021w = extrude(sketch021w, length = 1)
const sketch022w = startSketchOn(plane002)
|> startProfileAt([depth - 1, 74.1], %)
|> angledLine(angle = 180 - 23, endAbsoluteY = railHeight * 1.75 + 1)
|> xLine(length = -2.56)
|> angledLine(angle = -23, endAbsoluteX = depth - 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude022w = extrude(sketch022w, length = 1)
const sketch023w = startSketchOn(plane002)
|> startProfileAt([1, 69.53 + 1.75], %)
|> angledLine(angle = 90 - 23, endAbsoluteY = railHeight * 1.75 + 1)
|> xLine(length = 1.086)
|> angledLine(angle = 90 - 23, endAbsoluteX = 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude023w = extrude(sketch023w, length = 1)
const sketch024w = startSketchOn(plane002)
|> startProfileAt([1, 16.5], %)
|> angledLine(angle = -23, endAbsoluteY = 1)
|> xLine(length = -2.56)
|> angledLine(angle = -23, endAbsoluteX = 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude024w = extrude(sketch024w, length = 1)
const sketch025w = startSketchOn(plane002)
|> startProfileAt([1, 4], %)
|> angledLine(angle = -23, endAbsoluteY = 1)
|> xLine(length = -2.56)
|> angledLine(angle = -23, endAbsoluteX = 1)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude025w = extrude(sketch025w, length = 1)
// Base plane of rack
const plane003 = {
origin: [0, 0, 4.114],
x_axis: [1, 0, 0],
y_axis: [0, 1, 0],
z_axis: [0, 0, 1]
}
const sketch005 = startSketchOn(plane003)
|> startProfileAt([width - 1, 0], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [-width + 2, 3])
|> line(end = [0, 1])
|> line(end = [width - 2, -3])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude005 = extrude(sketch005, length = 1)
const sketch006 = startSketchOn(plane003)
|> startProfileAt([width - 1, 0 + 2.756], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [-width + 2, 3])
|> line(end = [0, 1])
|> line(end = [width - 2, -3])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude006 = extrude(sketch006, length = 1)
const sketch007 = startSketchOn(plane003)
|> startProfileAt([width - 1, depth], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [-width + 1, -depth + serverDepth + 3.8])
|> line(end = [1, 0])
|> line(end = [0, -.75])
|> line([
width - 2,
depth - serverDepth - 5 + .6
], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude007 = extrude(sketch007, length = 1)
const sketch008 = startSketchOn(plane003)
|> startProfileAt([width - 1, depth - 2.8], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [-width + 2, -depth + serverDepth + 4.4])
|> line(end = [0, -1.32])
|> line(end = [width - 2, depth - serverDepth - 4.4])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude008 = extrude(sketch008, length = 1)
// Top Plane
const plane004 = {
origin: [0, 0, 4.114 + railHeight * 1.75 + 2],
x_axis: [1, 0, 0],
y_axis: [0, 1, 0],
z_axis: [0, 0, 1]
}
const sketch005t = startSketchOn(plane004)
|> startProfileAt([width - 1, 0], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [-width + 2, 3])
|> line(end = [0, 1])
|> line(end = [width - 2, -3])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude005t = extrude(sketch005t, length = -1)
const sketch007t = startSketchOn(plane004)
|> startProfileAt([width - 1, depth], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [-width + 1, -depth + serverDepth + 3.8])
|> line(end = [1, 0])
|> line(end = [0, -.75])
|> line([
width - 2,
depth - serverDepth - 5 + .6
], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude007t = extrude(sketch007t, length = -1)
const sketch008t = startSketchOn(plane004)
|> startProfileAt([width - 1, 0 + 2.756], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [-width + 2, 3])
|> line(end = [0, 1])
|> line(end = [width - 2, -3])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude008t = extrude(sketch008t, length = -1)
const sketch009t = startSketchOn(plane004)
|> startProfileAt([width - 1, depth - 2.8], %)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [-width + 2, -depth + serverDepth + 4.4])
|> line(end = [0, -1.32])
|> line(end = [width - 2, depth - serverDepth - 4.4])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude009t = extrude(sketch009t, length = -1)
// front left vertical rail
// Define planes so the rail can be moved
const planeXYfl = {
origin: [0, 0, originStart[2]],
xAxis: [1.0, 0.0, 0.0],
yAxis: [0.0, 1.0, 0.0],
zAxis: [0.0, 0.0, 1.0]
}
const planeXZfl = {
origin: [originStart[1], originStart[0], originStart[2]],
xAxis: [0, 1, 0],
yAxis: [0, 0, 1],
zAxis: [1, 0, 0]
}
const planeYZfl = {
origin: [originStart[1], originStart[0], originStart[2]],
xAxis: [-1, 0, 0],
yAxis: [0, 0, 1],
zAxis: [1, 0, 0]
}
// Base flange
const sketch001fl = startSketchOn(planeXZfl)
|> startProfileAt([bendRad + thickness, 0], %)
|> angledLine(angle = 0, length = 2 - bendRad - thickness, tag = $rectangleSegmentA001fl)
|> angledLine(angle = segAng(rectangleSegmentA001fl) + 90, length = railHeight * 1.75, tag = $rectangleSegmentB001fl)
|> angledLine(angle = segAng(rectangleSegmentA001fl), length = -segLen(rectangleSegmentA001fl), tag = $rectangleSegmentC001fl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude001fl = extrude(sketch001fl, length = thickness)
// First flange
const sketch002fl = startSketchOn(planeYZfl)
|> startProfileAt([-bendRad - thickness, 0], %)
|> angledLine(angle = 180, length = 0.89 - bendRad - thickness, tag = $rectangleSegmentA002fl)
|> angledLine(angle = segAng(rectangleSegmentA002fl) - 90, length = railHeight * 1.75, tag = $rectangleSegmentB002fl)
|> angledLine(angle = segAng(rectangleSegmentA002fl), length = -segLen(rectangleSegmentA002fl), tag = $rectangleSegmentC002fl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude002fl = extrude(sketch002fl, length = thickness)
// Bend
const sketch003fl = startSketchOn(planeXYfl)
|> startProfileAt([0 + thickness + bendRad+originStart[1], originStart[0]], %)
2025-04-18 17:40:44 -05:00
|> arc(angleStart = 270, angleEnd = 180, radius = bendRad + thickness)
|> xLine(length = thickness)
2025-04-18 17:40:44 -05:00
|> arc(angleStart = 180, angleEnd = 270, radius = bendRad)
|> yLine(length = -thickness)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude003fl = extrude(sketch003fl, length = railHeight * 1.75)
// EIA-310-D standard hole pattern
const sketch004fl = startSketchOn(extrude002fl, face = 'START')
|> startProfileAt([0.72 + originStart[1], originStart[2] + 0.15], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA003fl)
|> angledLine(angle = segAng(rectangleSegmentA003fl) - 90, length = 0.375, tag = $rectangleSegmentB003fl)
|> angledLine(angle = segAng(rectangleSegmentA003fl), length = -segLen(rectangleSegmentA003fl), tag = $rectangleSegmentC003fl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude004fl = extrude(sketch004fl, length = -thickness)
// EIA-310-D standard hole pattern
const sketch005fl = startSketchOn(extrude002fl, face = 'START')
|> startProfileAt([0.72 + originStart[1],originStart[2] + 0.15 + .62], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA004fl)
|> angledLine(angle = segAng(rectangleSegmentA004fl) - 90, length = 0.375, tag = $rectangleSegmentB004fl)
|> angledLine(angle = segAng(rectangleSegmentA004fl), length = -segLen(rectangleSegmentA004fl), tag = $rectangleSegmentC004fl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude005fl = extrude(sketch005fl, length = -thickness)
// EIA-310-D standard hole pattern
const sketch006fl = startSketchOn(extrude002fl, face = 'START')
|> startProfileAt([
0.72 + originStart[1],
originStart[2] + 0.15 + 0.62 + 0.50
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA005fl)
|> angledLine(angle = segAng(rectangleSegmentA005fl) - 90, length = 0.375, tag = $rectangleSegmentB005fl)
|> angledLine(angle = segAng(rectangleSegmentA005fl), length = -segLen(rectangleSegmentA005fl), tag = $rectangleSegmentC005fl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude006fl = extrude(sketch006fl, length = -thickness)
// EIA-310-D standard hole pattern
const sketch007fl = startSketchOn(extrude001fl, face = 'START')
|> startProfileAt([-1.45 - originStart[0], originStart[2] + 0.15], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA006fl)
|> angledLine(angle = segAng(rectangleSegmentA006fl) - 90, length = 0.375, tag = $rectangleSegmentB006fl)
|> angledLine(angle = segAng(rectangleSegmentA006fl), length = -segLen(rectangleSegmentA006fl), tag = $rectangleSegmentC006fl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude007fl = extrude(sketch007fl, length = -thickness)
// EIA-310-D standard hole pattern
const sketch008fl = startSketchOn(extrude001fl, face = 'START')
|> startProfileAt([-1.45 - originStart[0],originStart[2] + 0.15 + 0.62], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA007fl)
|> angledLine(angle = segAng(rectangleSegmentA007fl) - 90, length = 0.375, tag = $rectangleSegmentB007fl)
|> angledLine(angle = segAng(rectangleSegmentA007fl), length = -segLen(rectangleSegmentA007fl), tag = $rectangleSegmentC007fl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude008fl = extrude(sketch008fl, length = -thickness)
// EIA-310-D standard hole pattern
const sketch009fl = startSketchOn(extrude001fl, face = 'START')
|> startProfileAt([
-1.45 - originStart[0],
originStart[2] + 0.15 + 0.62 + 0.5
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA008fl)
|> angledLine(angle = segAng(rectangleSegmentA008fl) - 90, length = 0.375, tag = $rectangleSegmentB008fl)
|> angledLine(angle = segAng(rectangleSegmentA008fl), length = -segLen(rectangleSegmentA008fl), tag = $rectangleSegmentC008fl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude009fl = extrude(sketch009fl, length = -thickness)
// define slots
const sketch010fl = startSketchOn(extrude001fl, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0],
originStart[2] + .81 - (.438 / 2)
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [-0.66 - originStart[0],originStart[2] + .81 + .438 / 2])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = 2,
distance = 1.22
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude010fl = extrude(sketch010fl, length = -thickness)
// define slots
const sketch011fl = startSketchOn(extrude001fl, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0],
originStart[2] + railHeight * 1.75 / 2 + .438 / 2
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0],
originStart[2] + railHeight * 1.75 / 2 - (.438 / 2)
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude011fl = extrude(sketch011fl, length = -thickness)
// define slots
const sketch012fl = startSketchOn(extrude001fl, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0], originStart[2] +
railHeight * 1.75 - .81 + .438 / 2
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0], originStart[2]+
railHeight * 1.75 - .81 - (.438 / 2)
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
|> patternLinear2d(
axis = [0, -1],
instances = 2,
distance = 1.22
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude012fl = extrude(sketch012fl, length = -thickness)
// FRONT RIGHT VERTICAL RAIL
// Define planes so the rail can be moved
const planeXYfr = {
origin: [0, 0, originStart[2]],
xAxis: [1.0, 0.0, 0.0],
yAxis: [0.0, 1.0, 0.0],
zAxis: [0.0, 0.0, 1.0]
}
const planeXZfr = {
origin: [
originStart[1] + width - thickness - 2,
originStart[0],
originStart[2]
],
xAxis: [0, 1, 0],
yAxis: [0, 0, 1],
zAxis: [1, 0, 0]
}
const planeYZfr = {
origin: [
originStart[1] + width - 2,
originStart[0],
originStart[2]
],
xAxis: [-1, 0, 0],
yAxis: [0, 0, 1],
zAxis: [1, 0, 0]
}
// Base flange
const sketch001fr = startSketchOn(planeXZfr)
|> startProfileAt([bendRad + thickness, 0], %)
|> angledLine(angle = 0, length = 2 - bendRad - thickness, tag = $rectangleSegmentA001fr)
|> angledLine(angle = segAng(rectangleSegmentA001fr) + 90, length = railHeight * 1.75, tag = $rectangleSegmentB001fr)
|> angledLine(angle = segAng(rectangleSegmentA001fr), length = -segLen(rectangleSegmentA001fr), tag = $rectangleSegmentC001fr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude001fr = extrude(sketch001fr, length = thickness)
// First flange
const sketch002fr = startSketchOn(planeYZfr)
|> startProfileAt([0.89, 0], %)
|> angledLine(angle = 180, length = 0.89 - bendRad - thickness, tag = $rectangleSegmentA002fr)
|> angledLine(angle = segAng(rectangleSegmentA002fr) - 90, length = railHeight * 1.75, tag = $rectangleSegmentB002fr)
|> angledLine(angle = segAng(rectangleSegmentA002fr), length = -segLen(rectangleSegmentA002fr), tag = $rectangleSegmentC002fr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude002fr = extrude(sketch002fr, length = thickness)
// Bend
const sketch003fr = startSketchOn(planeXYfr)
|> startProfileAt([
bendRad + originStart[1] + width - 2 - thickness - bendRad,
originStart[0] + bendRad + thickness
], %)
2025-04-18 17:40:44 -05:00
|> arc(angleStart = 0, angleEnd = -90, radius = bendRad)
|> yLine(length = -thickness)
2025-04-18 17:40:44 -05:00
|> arc(angleStart = -90, angleEnd = 0, radius = bendRad + thickness)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude003fr = extrude(sketch003fr, length = railHeight * 1.75)
// EIA-310-D standard hole pattern
const sketch004fr = startSketchOn(extrude002fr, face = 'START')
|> startProfileAt([
-0.35 + originStart[1] + width - 2,
originStart[2] + 0.15
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA003fr)
|> angledLine(angle = segAng(rectangleSegmentA003fr) - 90, length = 0.375, tag = $rectangleSegmentB003fr)
|> angledLine(angle = segAng(rectangleSegmentA003fr), length = -segLen(rectangleSegmentA003fr), tag = $rectangleSegmentC003fr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude004fr = extrude(sketch004fr, length = -thickness)
// EIA-310-D standard hole pattern
const sketch005fr = startSketchOn(extrude002fr, face = 'START')
|> startProfileAt([
-0.35 + originStart[1] + width - 2,
originStart[2] + 0.15 + .62
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA004fr)
|> angledLine(angle = segAng(rectangleSegmentA004fr) - 90, length = 0.375, tag = $rectangleSegmentB004fr)
|> angledLine(angle = segAng(rectangleSegmentA004fr), length = -segLen(rectangleSegmentA004fr), tag = $rectangleSegmentC004fr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude005fr = extrude(sketch005fr, length = -thickness)
// EIA-310-D standard hole pattern
const sketch006fr = startSketchOn(extrude002fr, face = 'START')
|> startProfileAt([
-0.35 + originStart[1] + width - 2,
originStart[2] + 0.15 + 0.62 + 0.50
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA005fr)
|> angledLine(angle = segAng(rectangleSegmentA005fr) - 90, length = 0.375, tag = $rectangleSegmentB005fr)
|> angledLine(angle = segAng(rectangleSegmentA005fr), length = -segLen(rectangleSegmentA005fr), tag = $rectangleSegmentC005fr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude006fr = extrude(sketch006fr, length = -thickness)
// EIA-310-D standard hole pattern
const sketch007fr = startSketchOn(extrude001fr, face = 'START')
|> startProfileAt([
-1.45 - originStart[0],
originStart[2] + 0.15
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA006fr)
|> angledLine(angle = segAng(rectangleSegmentA006fr) - 90, length = 0.375, tag = $rectangleSegmentB006fr)
|> angledLine(angle = segAng(rectangleSegmentA006fr), length = -segLen(rectangleSegmentA006fr), tag = $rectangleSegmentC006fr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude007fr = extrude(sketch007fr, length = -thickness)
// EIA-310-D standard hole pattern
const sketch008fr = startSketchOn(extrude001fr, face = 'START')
|> startProfileAt([
-1.45 - originStart[0],
originStart[2] + 0.15 + 0.62
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA007fr)
|> angledLine(angle = segAng(rectangleSegmentA007fr) - 90, length = 0.375, tag = $rectangleSegmentB007fr)
|> angledLine(angle = segAng(rectangleSegmentA007fr), length = -segLen(rectangleSegmentA007fr), tag = $rectangleSegmentC007fr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude008fr = extrude(sketch008fr, length = -thickness)
// EIA-310-D standard hole pattern
const sketch009fr = startSketchOn(extrude001fr, face = 'START')
|> startProfileAt([
-1.45 - originStart[0],
originStart[2] + 0.15 + 0.62 + 0.5
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA008fr)
|> angledLine(angle = segAng(rectangleSegmentA008fr) - 90, length = 0.375, tag = $rectangleSegmentB008fr)
|> angledLine(angle = segAng(rectangleSegmentA008fr), length = -segLen(rectangleSegmentA008fr), tag = $rectangleSegmentC008fr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude009fr = extrude(sketch009fr, length = -thickness)
// define slots
const sketch010fr = startSketchOn(extrude001fr, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0],
originStart[2] + .81 - (.438 / 2)
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0],
originStart[2] + .81 + .438 / 2
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = 2,
distance = 1.22
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude010fr = extrude(sketch010fr, length = -thickness)
// define slots
const sketch011fr = startSketchOn(extrude001fr, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0],
originStart[2] + railHeight * 1.75 / 2 + .438 / 2
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0],
originStart[2] + railHeight * 1.75 / 2 - (.438 / 2)
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude011fr = extrude(sketch011fr, length = -thickness)
// define slots
const sketch012fr = startSketchOn(extrude001fr, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0],
originStart[2] + railHeight * 1.75 - .81 + .438 / 2
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0],
originStart[2] + railHeight * 1.75 - .81 - (.438 / 2)
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
|> patternLinear2d(
axis = [0, -1],
instances = 2,
distance = 1.22
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude012fr = extrude(sketch012fr, length = -thickness)
// RIGHT REAR VERTICAL RAIL
// Define planes so the rail can be moved
const planeXYrr = {
origin: [0, serverDepth, originStart[2]],
xAxis: [1.0, 0.0, 0.0],
yAxis: [0.0, 1.0, 0.0],
zAxis: [0.0, 0.0, 1.0]
}
const planeXZrr = {
origin: [
originStart[1] + width - thickness - 2,
originStart[0]+serverDepth-2-bendRad,
originStart[2]
],
xAxis: [0, 1, 0],
yAxis: [0, 0, 1],
zAxis: [1, 0, 0]
}
const planeYZrr = {
origin: [
originStart[1] + width - 2,
originStart[0]+serverDepth,
originStart[2]
],
xAxis: [-1, 0, 0],
yAxis: [0, 0, 1],
zAxis: [1, 0, 0]
}
// Base flange
const sketch001rr = startSketchOn(planeXZrr)
|> startProfileAt([bendRad + thickness, 0], %)
|> angledLine(angle = 0, length = 2 - bendRad - thickness, tag = $rectangleSegmentA001rr)
|> angledLine(angle = segAng(rectangleSegmentA001rr) + 90, length = railHeight * 1.75, tag = $rectangleSegmentB001rr)
|> angledLine(angle = segAng(rectangleSegmentA001rr), length = -segLen(rectangleSegmentA001rr), tag = $rectangleSegmentC001rr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude001rr = extrude(sketch001rr, length = thickness)
// First flange
const sketch002rr = startSketchOn(planeYZrr)
|> startProfileAt([0.89, 0], %)
|> angledLine(angle = 180, length = 0.89 - bendRad - thickness, tag = $rectangleSegmentA002rr)
|> angledLine(angle = segAng(rectangleSegmentA002rr) - 90, length = railHeight * 1.75, tag = $rectangleSegmentB002rr)
|> angledLine(angle = segAng(rectangleSegmentA002rr), length = -segLen(rectangleSegmentA002rr), tag = $rectangleSegmentC002rr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude002rr = extrude(sketch002rr, length = thickness)
// Bend
const sketch003rr = startSketchOn(planeXYrr)
|> startProfileAt([
bendRad + originStart[1] + width - 2-bendRad,
originStart[0]-bendRad
], %)
2025-04-18 17:40:44 -05:00
|> arc(angleStart = 0, angleEnd = 90, radius = bendRad+thickness)
|> yLine(length = -thickness)
2025-04-18 17:40:44 -05:00
|> arc(angleStart = 90, angleEnd = 0, radius = bendRad)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude003rr = extrude(sketch003rr, length = railHeight * 1.75)
// EIA-310-D standard hole pattern
const sketch004rr = startSketchOn(extrude002rr, face = 'START')
|> startProfileAt([
-0.35 + originStart[1] + width - 2,
originStart[2] + 0.15
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA003rr)
|> angledLine(angle = segAng(rectangleSegmentA003rr) - 90, length = 0.375, tag = $rectangleSegmentB003rr)
|> angledLine(angle = segAng(rectangleSegmentA003rr), length = -segLen(rectangleSegmentA003rr), tag = $rectangleSegmentC003rr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude004rr = extrude(sketch004rr, length = -thickness)
// EIA-310-D standard hole pattern
const sketch005rr = startSketchOn(extrude002rr, face = 'START')
|> startProfileAt([
-0.35 + originStart[1] + width - 2,
originStart[2] + 0.15 + .62
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA004rr)
|> angledLine(angle = segAng(rectangleSegmentA004rr) - 90, length = 0.375, tag = $rectangleSegmentB004rr)
|> angledLine(angle = segAng(rectangleSegmentA004rr), length = -segLen(rectangleSegmentA004rr), tag = $rectangleSegmentC004rr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude005rr = extrude(sketch005rr, length = -thickness)
// EIA-310-D standard hole pattern
const sketch006rr = startSketchOn(extrude002rr, face = 'START')
|> startProfileAt([
-0.35 + originStart[1] + width - 2,
originStart[2] + 0.15 + 0.62 + 0.50
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA005rr)
|> angledLine(angle = segAng(rectangleSegmentA005rr) - 90, length = 0.375, tag = $rectangleSegmentB005rr)
|> angledLine(angle = segAng(rectangleSegmentA005rr), length = -segLen(rectangleSegmentA005rr), tag = $rectangleSegmentC005rr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude006rr = extrude(sketch006rr, length = -thickness)
// EIA-310-D standard hole pattern
const sketch007rr = startSketchOn(extrude001rr, face = 'START')
|> startProfileAt([
-originStart[0]-serverDepth+1.7,
originStart[2] + 0.15
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA006rr)
|> angledLine(angle = segAng(rectangleSegmentA006rr) - 90, length = 0.375, tag = $rectangleSegmentB006rr)
|> angledLine(angle = segAng(rectangleSegmentA006rr), length = -segLen(rectangleSegmentA006rr), tag = $rectangleSegmentC006rr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude007rr = extrude(sketch007rr, length = -thickness)
// EIA-310-D standard hole pattern
const sketch008rr = startSketchOn(extrude001rr, face = 'START')
|> startProfileAt([
-originStart[0]-serverDepth+1.7,
originStart[2] + 0.15 + 0.62
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA007rr)
|> angledLine(angle = segAng(rectangleSegmentA007rr) - 90, length = 0.375, tag = $rectangleSegmentB007rr)
|> angledLine(angle = segAng(rectangleSegmentA007rr), length = -segLen(rectangleSegmentA007rr), tag = $rectangleSegmentC007rr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude008rr = extrude(sketch008rr, length = -thickness)
// EIA-310-D standard hole pattern
const sketch009rr = startSketchOn(extrude001rr, face = 'START')
|> startProfileAt([
-originStart[0]-serverDepth+1.7,
originStart[2] + 0.15 + 0.62 + 0.5
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA008rr)
|> angledLine(angle = segAng(rectangleSegmentA008rr) - 90, length = 0.375, tag = $rectangleSegmentB008rr)
|> angledLine(angle = segAng(rectangleSegmentA008rr), length = -segLen(rectangleSegmentA008rr), tag = $rectangleSegmentC008rr)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude009rr = extrude(sketch009rr, length = -thickness)
// define slots
const sketch010rr = startSketchOn(extrude001rr, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0]+1.5-serverDepth,
originStart[2] + .81 - (.438 / 2)
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0]+1.5-serverDepth,
originStart[2] + .81 + .438 / 2
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = 2,
distance = 1.22
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude010rr = extrude(sketch010rr, length = -thickness)
// define slots
const sketch011rr = startSketchOn(extrude001rr, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0]+1.5-serverDepth,
originStart[2] + railHeight * 1.75 / 2 + .438 / 2
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0]+1.5-serverDepth,
originStart[2] + railHeight * 1.75 / 2 - (.438 / 2)
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude011rr = extrude(sketch011rr, length = -thickness)
// define slots
const sketch012rr = startSketchOn(extrude001rr, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0]+1.5-serverDepth,
originStart[2] + railHeight * 1.75 - .81 + .438 / 2
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0]+1.5-serverDepth,
originStart[2] + railHeight * 1.75 - .81 - (.438 / 2)
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
|> patternLinear2d(
axis = [0, -1],
instances = 2,
distance = 1.22
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude012rr = extrude(sketch012rr, length = -thickness)
// REAR LEFT VETCIAL RAIL
// Define planes so the rail can be moved
const planeXYrl = {
origin: [0, serverDepth, originStart[2]],
xAxis: [1.0, 0.0, 0.0],
yAxis: [0.0, 1.0, 0.0],
zAxis: [0.0, 0.0, 1.0]
}
const planeXZrl = {
origin: [
originStart[1],
originStart[0] + serverDepth - 2 - bendRad,
originStart[2]
],
xAxis: [0, 1, 0],
yAxis: [0, 0, 1],
zAxis: [1, 0, 0]
}
const planeYZrl = {
origin: [
originStart[1],
originStart[0] + serverDepth,
originStart[2]
],
xAxis: [-1, 0, 0],
yAxis: [0, 0, 1],
zAxis: [1, 0, 0]
}
// Base flange
const sketch001rl = startSketchOn(planeXZrl)
|> startProfileAt([bendRad + thickness, 0], %)
|> angledLine(angle = 0, length = 2 - bendRad - thickness, tag = $rectangleSegmentA001rl)
|> angledLine(angle = segAng(rectangleSegmentA001rl) + 90, length = railHeight * 1.75, tag = $rectangleSegmentB001rl)
|> angledLine(angle = segAng(rectangleSegmentA001rl), length = -segLen(rectangleSegmentA001rl), tag = $rectangleSegmentC001rl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude001rl = extrude(sketch001rl, length = thickness)
// First flange
const sketch002rl = startSketchOn(planeYZrl)
|> startProfileAt([-bendRad - thickness, 0], %)
|> angledLine(angle = 180, length = 0.89 - bendRad - thickness, tag = $rectangleSegmentA002rl)
|> angledLine(angle = segAng(rectangleSegmentA002rl) - 90, length = railHeight * 1.75, tag = $rectangleSegmentB002rl)
|> angledLine(angle = segAng(rectangleSegmentA002rl), length = -segLen(rectangleSegmentA002rl), tag = $rectangleSegmentC002rl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude002rl = extrude(sketch002rl, length = thickness)
// Bend
const sketch003rl = startSketchOn(planeXYrl)
|> startProfileAt([
bendRad + originStart[1] + thickness,
originStart[0]
], %)
2025-04-18 17:40:44 -05:00
|> arc(angleStart = 90, angleEnd = 180, radius = bendRad)
|> xLine(length = -thickness)
2025-04-18 17:40:44 -05:00
|> arc(angleStart = 180, angleEnd = 90, radius = bendRad + thickness)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude003rl = extrude(sketch003rl, length = railHeight * 1.75)
// EIA-310-D standard hole pattern
const sketch004rl = startSketchOn(extrude002rl, face = 'START')
|> startProfileAt([
-1.38 + 2 + thickness + originStart[1],
originStart[2] + 0.15
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA003rl)
|> angledLine(angle = segAng(rectangleSegmentA003rl) - 90, length = 0.375, tag = $rectangleSegmentB003rl)
|> angledLine(angle = segAng(rectangleSegmentA003rl), length = -segLen(rectangleSegmentA003rl), tag = $rectangleSegmentC003rl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude004rl = extrude(sketch004rl, length = -thickness)
// EIA-310-D standard hole pattern
const sketch005rl = startSketchOn(extrude002rl, face = 'START')
|> startProfileAt([
-1.38 + 2 + thickness + originStart[1],
originStart[2] + 0.15 + .62
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA004rl)
|> angledLine(angle = segAng(rectangleSegmentA004rl) - 90, length = 0.375, tag = $rectangleSegmentB004rl)
|> angledLine(angle = segAng(rectangleSegmentA004rl), length = -segLen(rectangleSegmentA004rl), tag = $rectangleSegmentC004rl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude005rl = extrude(sketch005rl, length = -thickness)
// EIA-310-D standard hole pattern
const sketch006rl = startSketchOn(extrude002rl, face = 'START')
|> startProfileAt([
-1.38 + 2 + thickness + originStart[1],
originStart[2] + 0.15 + 0.62 + 0.50
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA005rl)
|> angledLine(angle = segAng(rectangleSegmentA005rl) - 90, length = 0.375, tag = $rectangleSegmentB005rl)
|> angledLine(angle = segAng(rectangleSegmentA005rl), length = -segLen(rectangleSegmentA005rl), tag = $rectangleSegmentC005rl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude006rl = extrude(sketch006rl, length = -thickness)
// EIA-310-D standard hole pattern
const sketch007rl = startSketchOn(extrude001rl, face = 'START')
|> startProfileAt([
1.75 - originStart[0] - serverDepth,
originStart[2] + 0.15
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA006rl)
|> angledLine(angle = segAng(rectangleSegmentA006rl) - 90, length = 0.375, tag = $rectangleSegmentB006rl)
|> angledLine(angle = segAng(rectangleSegmentA006rl), length = -segLen(rectangleSegmentA006rl), tag = $rectangleSegmentC006rl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude007rl = extrude(sketch007rl, length = -thickness)
// EIA-310-D standard hole pattern
const sketch008rl = startSketchOn(extrude001rl, face = 'START')
|> startProfileAt([
1.75 - originStart[0] - serverDepth,
originStart[2] + 0.15 + 0.62
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA007rl)
|> angledLine(angle = segAng(rectangleSegmentA007rl) - 90, length = 0.375, tag = $rectangleSegmentB007rl)
|> angledLine(angle = segAng(rectangleSegmentA007rl), length = -segLen(rectangleSegmentA007rl), tag = $rectangleSegmentC007rl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude008rl = extrude(sketch008rl, length = -thickness)
// EIA-310-D standard hole pattern
const sketch009rl = startSketchOn(extrude001rl, face = 'START')
|> startProfileAt([
1.75 - originStart[0] - serverDepth,
originStart[2] + 0.15 + 0.62 + 0.5
], %)
|> angledLine(angle = 0, length = -0.375, tag = $rectangleSegmentA008rl)
|> angledLine(angle = segAng(rectangleSegmentA008rl) - 90, length = 0.375, tag = $rectangleSegmentB008rl)
|> angledLine(angle = segAng(rectangleSegmentA008rl), length = -segLen(rectangleSegmentA008rl), tag = $rectangleSegmentC008rl)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = railHeight,
distance = 0.62 + 0.62 + 0.5
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude009rl = extrude(sketch009rl, length = -thickness)
// define slots
const sketch010rl = startSketchOn(extrude001rl, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0] - serverDepth + 1.5,
originStart[2] + .81 - (.438 / 2)
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0] - serverDepth + 1.5,
originStart[2] + .81 + .438 / 2
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
|> patternLinear2d(
axis = [0, 1],
instances = 2,
distance = 1.22
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude010rl = extrude(sketch010rl, length = -thickness)
// define slots
const sketch011rl = startSketchOn(extrude001rl, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0] - serverDepth + 1.5,
originStart[2] + railHeight * 1.75 / 2 + .438 / 2
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0] - serverDepth + 1.5,
originStart[2] + railHeight * 1.75 / 2 - (.438 / 2)
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude011rl = extrude(sketch011rl, length = -thickness)
// define slots
const sketch012rl = startSketchOn(extrude001rl, face = 'START')
|> startProfileAt([
-1.12 + (.75 - .438) / 2 - originStart[0] - serverDepth + 1.5,
originStart[2] + railHeight * 1.75 - .81 + .438 / 2
], %)
|> xLine(length = 0.75 - .438)
|> tangentialArc(endAbsolute = [
-0.66 - originStart[0] - serverDepth + 1.5,
originStart[2] + railHeight * 1.75 - .81 - (.438 / 2)
])
|> xLine(length = -0.75 + .438)
|> tangentialArc(endAbsolute = [profileStartX(%), profileStartY(%)])
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> close()
|> patternLinear2d(
axis = [0, -1],
instances = 2,
distance = 1.22
)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude012rl = extrude(sketch012rl, length = -thickness)
// GENERATE SERVER MODELS
// Define planes so the server can be moved
fn streamServer = (serverPos) => {
const planeXYs = {
origin: [0, 0 + 2, 4.114 + 1 + serverPos * 1.75],
xAxis: [1.0, 0.0, 0.0],
yAxis: [0.0, 1.0, 0.0],
zAxis: [0.0, 0.0, 1.0]
}
const planeXZs = {
origin: [0 + 1.2 + .1, 0 + 2, 0],
xAxis: [0, -1, 0],
yAxis: [0, 0, 1],
zAxis: [1, 0, 0]
}
const planeXZs2 = {
origin: [18 + 1.4, 0 + 2, 0],
xAxis: [0, -1, 0],
yAxis: [0, 0, 1],
zAxis: [1, 0, 0]
}
const sketch001s = startSketchOn(planeXYs)
|> startProfileAt([1 + 1.2, 1], %)
|> angledLine(angle = 0, length = 17.2, tag = $rectangleSegmentA001)
|> angledLine(angle = segAng(rectangleSegmentA001) + 90, length = 29, tag = $rectangleSegmentB001)
|> angledLine(angle = segAng(rectangleSegmentA001), length = -segLen(rectangleSegmentA001), tag = $rectangleSegmentC001)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
const extrude001s = extrude(sketch001s, length = 7)
const sketch002s = startSketchOn(planeXZs)
|> startProfileAt([-1, 4.114 + 1 + serverPos * 1.75], %)
|> yLine(length = 6.98)
|> xLine(length = 0.2)
|> yLine(length = -0.36)
|> xLine(length = 0.5)
|> tangentialArc(endAbsolute = [
0.3,
17.15 + 4.114 + 1 + serverPos * 1.75 - 11.114
])
|> yLine(length = -1.77)
|> tangentialArc(endAbsolute = [
-0.13,
14.89 + 4.114 + 1 + serverPos * 1.75 - 11.114
])
|> xLine(length = -0.52)
|> yLine(length = -0.42)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [0.34, -0.15])
|> yLine(length = -2.97)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude002s = extrude(sketch002s, length = 1.8 / 2)
const sketch003s = startSketchOn(planeXZs2)
|> startProfileAt([-1, 4.114 + 1 + serverPos * 1.75], %)
|> yLine(length = 6.98)
|> xLine(length = 0.2)
|> yLine(length = -0.36)
|> xLine(length = 0.5)
|> tangentialArc(endAbsolute = [
0.3,
17.15 + 4.114 + 1 + serverPos * 1.75 - 11.114
])
|> yLine(length = -1.77)
|> tangentialArc(endAbsolute = [
-0.13,
14.89 + 4.114 + 1 + serverPos * 1.75 - 11.114
])
|> xLine(length = -0.52)
|> yLine(length = -0.42)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(end = [0.34, -0.15])
|> yLine(length = -2.97)
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
KCL: Use keyword arguments for line, lineTo, extrude and close (#5249) Part of #4600. PR: https://github.com/KittyCAD/modeling-app/pull/4826 # Changes to KCL stdlib - `line(point, sketch, tag)` and `lineTo(point, sketch, tag)` are combined into `line(@sketch, end?, endAbsolute?, tag?)` - `close(sketch, tag?)` is now `close(@sketch, tag?)` - `extrude(length, sketch)` is now `extrude(@sketch, length)` Note that if a parameter starts with `@` like `@sketch`, it doesn't have any label when called, so you call it like this: ``` sketch = startSketchAt([0, 0]) line(sketch, end = [3, 3], tag = $hi) ``` Note also that if you're using a `|>` pipeline, you can omit the `@` argument and it will be assumed to be the LHS of the `|>`. So the above could be written as ``` sketch = startSketchAt([0, 0]) |> line(end = [3, 3], tag = $hi) ``` Also changes frontend tests to use KittyCAD/kcl-samples#139 instead of its main The regex find-and-replace I use for migrating code (note these don't work with multi-line expressions) are: ``` line\(([^=]*), %\) line(end = $1) line\((.*), %, (.*)\) line(end = $1, tag = $2) lineTo\((.*), %\) line(endAbsolute = $1) lineTo\((.*), %, (.*)\) line(endAbsolute = $1, tag = $2) extrude\((.*), %\) extrude(length = $1) extrude\(([^=]*), ([a-zA-Z0-9]+)\) extrude($2, length = $1) close\(%, (.*)\) close(tag = $1) ``` # Selected notes from commits before I squash them all * Fix test 'yRelative to horizontal distance' Fixes: - Make a lineTo helper - Fix pathToNode to go through the labeled arg .arg property * Fix test by changing lookups into transformMap Parts of the code assumed that `line` is always a relative call. But actually now it might be absolute, if it's got an `endAbsolute` parameter. So, change whether to look up `line` or `lineTo` and the relevant absolute or relative line types based on that parameter. * Stop asserting on exact source ranges When I changed line to kwargs, all the source ranges we assert on became slightly different. I find these assertions to be very very low value. So I'm removing them. * Fix more tests: getConstraintType calls weren't checking if the 'line' fn was absolute or relative. * Fixed another queryAst test There were 2 problems: - Test was looking for the old style of `line` call to choose an offset for pathToNode - Test assumed that the `tag` param was always the third one, but in a kwarg call, you have to look it up by label * Fix test: traverse was not handling CallExpressionKw * Fix another test, addTagKw addTag helper was not aware of kw args. * Convert close from positional to kwargs If the close() call has 0 args, or a single unlabeled arg, the parser interprets it as a CallExpression (positional) not a CallExpressionKw. But then if a codemod wants to add a tag to it, it tries adding a kwarg called 'tag', which fails because the CallExpression doesn't need kwargs inserted into it. The fix is: change the node from CallExpression to CallExpressionKw, and update getNodeFromPath to take a 'replacement' arg, so we can replace the old node with the new node in the AST. * Fix the last test Test was looking for `lineTo` as a substring of the input KCL program. But there's no more lineTo function, so I changed it to look for line() with an endAbsolute arg, which is the new equivalent. Also changed the getConstraintInfo code to look up the lineTo if using line with endAbsolute. * Fix many bad regex find-replaces I wrote a regex find-and-replace which converted `line` calls from positional to keyword calls. But it was accidentally applied to more places than it should be, for example, angledLine, xLine and yLine calls. Fixes this. * Fixes test 'Basic sketch › code pane closed at start' Problem was, the getNodeFromPath call might not actually find a callExpressionKw, it might find a callExpression. So the `giveSketchFnCallTag` thought it was modifying a kwargs call, but it was actually modifying a positional call. This meant it tried to push a labeled argument in, rather than a normal arg, and a lot of other problems. Fixed by doing runtime typechecking. * Fix: Optional args given with wrong type were silently ignored Optional args don't have to be given. But if the user gives them, they should be the right type. Bug: if the KCL interpreter found an optional arg, which was given, but was the wrong type, it would ignore it and pretend the arg was never given at all. This was confusing for users. Fix: Now if you give an optional arg, but it's the wrong type, KCL will emit a type error just like it would for a mandatory argument. --------- Signed-off-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: Nick Cameron <nrc@ncameron.org> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@kittycad.io> Co-authored-by: Kevin Nadro <kevin@zoo.dev> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-04 08:31:43 -06:00
const extrude003s = extrude(sketch003s, length = 1.8 / 2)
return extrude003s
}
// generate servers at Z height in U's
// streamServer(4)
// streamServer(8)
// streamServer(12)
// streamServer(16)
// streamServer(20)
// streamServer(29)
// streamServer(33)
// streamServer(37)