Files
modeling-app/src/lang/modifyAst/addEdgeTreatment.test.ts

1019 lines
34 KiB
TypeScript
Raw Normal View History

import {
assertParse,
recast,
initPromise,
PathToNode,
Program,
CallExpression,
makeDefaultPlanes,
PipeExpression,
VariableDeclarator,
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
SourceRange,
topLevelRange,
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
CallExpressionKw,
} from '../wasm'
import {
EdgeTreatmentType,
getPathToExtrudeForSegmentSelection,
hasValidEdgeTreatmentSelection,
isTagUsedInEdgeTreatment,
modifyAstWithEdgeTreatmentAndTag,
FilletParameters,
ChamferParameters,
EdgeTreatmentParameters,
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
deleteEdgeTreatment,
} from './addEdgeTreatment'
import { getNodeFromPath } from '../queryAst'
import { getNodePathFromSourceRange } from 'lang/queryAstNodePathUtils'
import { createLiteral } from 'lang/modifyAst'
import { err } from 'lib/trap'
Point-and-click Shell (#4666) * WIP: experimenting with Loft UI Relates to #4470 * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Add selection guard * Working loft for two sketches in the right hardcoded order * First pass at handling more than 2 sketches * WIP selections * WIP selections * More checks * Appends the loft line after the 'last' sketch in the code * Clean up * Enable multiple selections after the button click * First point-click loft test (not working locally, loft gets inserted at the wrong place) * Lint * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * Clean up and working pw test * Add test for doesSceneHaveSweepableSketch with count = 2 * Clean up loftSketches function * Add pw test for preselected sketches * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Move to fromPromise-based Actor * Move error logic out of loftSketches, fix pw tests * Remove comments * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Fix typo * Revert snapshots * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger CI * WIP: initial shell code addition * Rollback pw values to pre cam change * WIP: more additions * WIP: closer * WIP: first time working shell mod * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * Add extrude lookup for more generic shell * Handle walls * Add pw tests for cap shell * Add shell wall test * Fix lint * Add selection guard and clean up * Lint fix * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * WIP mutliple faces * WIP circular dep * Lint * Look at this (photo)Graph *in the voice of Nickelback* * Trigger CI * Working multi-face shell across types * Cap and wall pw test * Apply suggestions from Frank's review Co-authored-by: Frank Noirot <frank@zoo.dev> * Fix test annotations * Add unit tests for doesSceneHaveExtrudedSketch * Manual resolution of snapshot conflicts * Fix assertParse * Updated pathToNode construct --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Frank Noirot <frank@zoo.dev>
2024-12-09 15:20:48 -05:00
import { Selection, Selections } from 'lib/selections'
import {
codeManager,
editorManager,
engineCommandManager,
kclManager,
} from 'lib/singletons'
import { VITE_KC_DEV_TOKEN } from 'env'
Selections Refactor (#4381) * selection stuff * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest) * trigger CI * fix bugs * some edge cut stuff * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * fix sketch mode issues * fix more tests, selection in sketch related * more test fixing * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger ci * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger ci * more sketch mode selection fixes * fix unit tests * rename function * remove .only * migrate a more selections types * migrate a more selections types * migrate a more selections types * lint * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * fix bad pathToNode issue * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * fix sketch on face * migrate a more selections types * migrate a more selections types * fix code selection of fillets * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * fix bad path to node, looks like a race * migrate a more selections types * migrate a more selections types * fix cmd bar selections * fix cmd bar selections * fix display issues * migrate a more selections types * Revert "migrate a more selections types" This reverts commit 0d0e453bbbb82d10472b6c9e753c53c8b7e3609d. * migrate a more selections types * clean up1 * clean up 2 * fix types after main merge * review tweaks * fix wall selection bug * Update src/lang/std/engineConnection.ts Co-authored-by: Jonathan Tran <jonnytran@gmail.com> * add franks TODO comment * fix type after main merge, plus a touch of clean up --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2024-11-21 15:04:30 +11:00
import { isOverlap } from 'lib/utils'
import { codeRefFromRange } from 'lang/std/artifactGraph'
beforeAll(async () => {
await initPromise
// THESE TEST WILL FAIL without VITE_KC_DEV_TOKEN set in .env.development.local
await new Promise((resolve) => {
engineCommandManager.start({
token: VITE_KC_DEV_TOKEN,
width: 256,
height: 256,
makeDefaultPlanes: () => makeDefaultPlanes(engineCommandManager),
setMediaStream: () => {},
setIsStreamReady: () => {},
callbackOnEngineLiteConnect: () => {
resolve(true)
},
})
})
}, 30_000)
afterAll(() => {
engineCommandManager.tearDown()
})
const dependencies = {
kclManager,
engineCommandManager,
editorManager,
codeManager,
}
const runGetPathToExtrudeForSegmentSelectionTest = async (
code: string,
selectedSegmentSnippet: string,
expectedExtrudeSnippet: string
) => {
// helpers
function getExtrudeExpression(
ast: Program,
pathToExtrudeNode: PathToNode
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
): CallExpression | CallExpressionKw | PipeExpression | undefined | Error {
if (pathToExtrudeNode.length === 0) return undefined // no extrude node
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 extrudeNodeResult = getNodeFromPath<
CallExpression | CallExpressionKw
>(ast, pathToExtrudeNode)
if (err(extrudeNodeResult)) {
return extrudeNodeResult
}
return extrudeNodeResult.node
}
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
function getExpectedExtrudeExpression(
ast: Program,
code: string,
expectedExtrudeSnippet: string
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
): CallExpression | CallExpressionKw | PipeExpression | Error {
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
const extrudeRange = topLevelRange(
code.indexOf(expectedExtrudeSnippet),
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
code.indexOf(expectedExtrudeSnippet) + expectedExtrudeSnippet.length
)
const expectedExtrudePath = getNodePathFromSourceRange(ast, extrudeRange)
const expectedExtrudeNodeResult = getNodeFromPath<
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
VariableDeclarator | CallExpression | CallExpressionKw
>(ast, expectedExtrudePath)
if (err(expectedExtrudeNodeResult)) {
return expectedExtrudeNodeResult
}
const expectedExtrudeNode = expectedExtrudeNodeResult.node
// check whether extrude is in the sketch pipe
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 extrudeInSketchPipe =
expectedExtrudeNode.type === 'CallExpression' ||
expectedExtrudeNode.type === 'CallExpressionKw'
if (extrudeInSketchPipe) {
return expectedExtrudeNode
}
if (!extrudeInSketchPipe) {
const init = expectedExtrudeNode.init
if (
init.type !== 'CallExpression' &&
init.type !== 'CallExpressionKw' &&
init.type !== 'PipeExpression'
) {
return new Error(
'Expected extrude expression is not a CallExpression or PipeExpression'
)
}
return init
}
return new Error('Expected extrude expression not found')
}
// ast
const ast = assertParse(code)
// range
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
const segmentRange = topLevelRange(
code.indexOf(selectedSegmentSnippet),
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
code.indexOf(selectedSegmentSnippet) + selectedSegmentSnippet.length
)
// executeAst and artifactGraph
await kclManager.executeAst({ ast })
const artifactGraph = engineCommandManager.artifactGraph
// find artifact
const maybeArtifact = [...artifactGraph].find(([, artifact]) => {
if (!('codeRef' in artifact && artifact.codeRef)) return false
return isOverlap(artifact.codeRef.range, segmentRange)
})
// build selection
const selection: Selection = {
codeRef: codeRefFromRange(segmentRange, ast),
artifact: maybeArtifact ? maybeArtifact[1] : undefined,
}
// get extrude expression
const pathResult = getPathToExtrudeForSegmentSelection(
ast,
selection,
artifactGraph
)
if (err(pathResult)) return pathResult
const { pathToExtrudeNode } = pathResult
const extrudeExpression = getExtrudeExpression(ast, pathToExtrudeNode)
// test
if (expectedExtrudeSnippet) {
const expectedExtrudeExpression = getExpectedExtrudeExpression(
ast,
code,
expectedExtrudeSnippet
)
if (err(expectedExtrudeExpression)) return expectedExtrudeExpression
expect(extrudeExpression).toEqual(expectedExtrudeExpression)
} else {
expect(extrudeExpression).toBeUndefined()
}
}
describe('Testing getPathToExtrudeForSegmentSelection', () => {
it('should return the correct paths for a valid selection and extrusion', async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)`
const selectedSegmentSnippet = `line(end = [20, 0])`
const expectedExtrudeSnippet = `extrude001 = extrude(sketch001, length = -15)`
await runGetPathToExtrudeForSegmentSelectionTest(
code,
selectedSegmentSnippet,
expectedExtrudeSnippet
)
}, 5_000)
it('should return the correct paths when extrusion occurs within the sketch pipe', async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> extrude(length = 15)`
const selectedSegmentSnippet = `line(end = [20, 0])`
const expectedExtrudeSnippet = `extrude(length = 15)`
await runGetPathToExtrudeForSegmentSelectionTest(
code,
selectedSegmentSnippet,
expectedExtrudeSnippet
)
}, 5_000)
it('should return the correct paths for a valid selection and extrusion in case of several extrusions and sketches', async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-30, 30], %)
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 = [15, 0])
|> line(end = [0, -15])
|> line(end = [-15, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
sketch002 = startSketchOn('XY')
|> startProfileAt([30, 30], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
sketch003 = startSketchOn('XY')
|> startProfileAt([30, -30], %)
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 = [25, 0])
|> line(end = [0, -25])
|> line(end = [-25, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
extrude002 = extrude(sketch002, length = -15)
extrude003 = extrude(sketch003, length = -15)`
const selectedSegmentSnippet = `line(end = [20, 0])`
const expectedExtrudeSnippet = `extrude002 = extrude(sketch002, length = -15)`
await runGetPathToExtrudeForSegmentSelectionTest(
code,
selectedSegmentSnippet,
expectedExtrudeSnippet
)
})
it('should return the correct paths for a (piped) extrude based on the other body (face)', async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-25, -25], %)
|> yLine(50, %)
|> xLine(50, %)
|> yLine(-50, %)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> extrude(length = 50)
sketch002 = startSketchOn(sketch001, 'END')
|> startProfileAt([-15, -15], %)
|> yLine(30, %)
|> xLine(30, %)
|> yLine(-30, %)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> extrude(length = 30)`
const selectedSegmentSnippet = `xLine(30, %)`
const expectedExtrudeSnippet = `extrude(length = 30)`
await runGetPathToExtrudeForSegmentSelectionTest(
code,
selectedSegmentSnippet,
expectedExtrudeSnippet
)
})
it('should return the correct paths for a (non-piped) extrude based on the other body (face)', async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-25, -25], %)
|> yLine(50, %)
|> xLine(50, %)
|> yLine(-50, %)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = 50)
sketch002 = startSketchOn(extrude001, 'END')
|> startProfileAt([-15, -15], %)
|> yLine(30, %)
|> xLine(30, %)
|> yLine(-30, %)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude002 = extrude(sketch002, length = 30)`
const selectedSegmentSnippet = `xLine(30, %)`
const expectedExtrudeSnippet = `extrude002 = extrude(sketch002, length = 30)`
await runGetPathToExtrudeForSegmentSelectionTest(
code,
selectedSegmentSnippet,
expectedExtrudeSnippet
)
})
it('should not return any path for missing extrusion', async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-30, 30], %)
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 = [15, 0])
|> line(end = [0, -15])
|> line(end = [-15, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
sketch002 = startSketchOn('XY')
|> startProfileAt([30, 30], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
sketch003 = startSketchOn('XY')
|> startProfileAt([30, -30], %)
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 = [25, 0])
|> line(end = [0, -25])
|> line(end = [-25, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
extrude003 = extrude(sketch003, length = -15)`
const selectedSegmentSnippet = `line(end = [20, 0])`
const expectedExtrudeSnippet = ``
await runGetPathToExtrudeForSegmentSelectionTest(
code,
selectedSegmentSnippet,
expectedExtrudeSnippet
)
}, 5_000)
})
const runModifyAstCloneWithEdgeTreatmentAndTag = async (
code: string,
selectionSnippets: Array<string>,
parameters: EdgeTreatmentParameters,
expectedCode: string
) => {
// ast
const ast = assertParse(code)
// selection
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
const segmentRanges: Array<SourceRange> = selectionSnippets.map(
(selectionSnippet) =>
topLevelRange(
code.indexOf(selectionSnippet),
code.indexOf(selectionSnippet) + selectionSnippet.length
)
)
// executeAst
await kclManager.executeAst({ ast })
Selections Refactor (#4381) * selection stuff * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest) * trigger CI * fix bugs * some edge cut stuff * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * fix sketch mode issues * fix more tests, selection in sketch related * more test fixing * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger ci * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger ci * more sketch mode selection fixes * fix unit tests * rename function * remove .only * migrate a more selections types * migrate a more selections types * migrate a more selections types * lint * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * fix bad pathToNode issue * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * fix sketch on face * migrate a more selections types * migrate a more selections types * fix code selection of fillets * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * fix bad path to node, looks like a race * migrate a more selections types * migrate a more selections types * fix cmd bar selections * fix cmd bar selections * fix display issues * migrate a more selections types * Revert "migrate a more selections types" This reverts commit 0d0e453bbbb82d10472b6c9e753c53c8b7e3609d. * migrate a more selections types * clean up1 * clean up 2 * fix types after main merge * review tweaks * fix wall selection bug * Update src/lang/std/engineConnection.ts Co-authored-by: Jonathan Tran <jonnytran@gmail.com> * add franks TODO comment * fix type after main merge, plus a touch of clean up --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2024-11-21 15:04:30 +11:00
const artifactGraph = engineCommandManager.artifactGraph
const selection: Selections = {
graphSelections: segmentRanges.map((segmentRange) => {
const maybeArtifact = [...artifactGraph].find(([, a]) => {
Add edit flows for extrude and offset plane operations (#5045) * Start implementing a "prepareToEdit" callback for extrude * Start of generic edit flow for operations * Actually invoking command bar send generically on double-click * Refactor: break out non-React hook helper to calculate Kcl expression value * Add unit tests, fmt * Integrate helper to get calculated KclExpression * Clean up unused imports, simplify use of `programMemoryFromVariables` * Implement basic extrude editing * Refactor: move DefaultPlanesStr to its own lib file * Add support for editing offset planes * Add Edit right-click menu option * Turn off edit flow for sketch for now * Add e2e tests for sketch and offset plane editing, fix bug found with offset plane editing * Add failing e2e extrude edit test * Remove action version of extrude AST mod * Fix behavior when adding a constant while editing operation, fixing e2e test * Patch in changes from 61b02b570394f11afbd04d0d126d87305165c73c * Remove shell's prepareToEdit * Add other Surface types to `artifactIsPlaneWithPaths` * refactor: rename `item` to `operation` * Allow `prepareToEdit` to fail with a toast, signal sketch-on-offset is unimplemented * Rework sketch e2e test to test several working and failing cases * Fix tsc errors related to making `codeRef` optional * Make basic error messages more friendly * fmt * Reset modifyAst.ts to main * Fix broken artifactGraph unit test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Remove unused import * Look at this (photo)Graph *in the voice of Nickelback* * Make the offset plane insert at the end, not one before * Fix bug caught by e2e test failure with "Command needs review" logic * Update src/machines/modelingMachine.ts Co-authored-by: Pierre Jacquier <pierrejacquier39@gmail.com> * Remove console logs per @pierremtb * Update src/components/CommandBar/CommandBarHeader.tsx Co-authored-by: Jonathan Tran <jonnytran@gmail.com> * Use better programMemory init thanks @jtran * Fix tsc post merge of #5068 * Fix logic for `artifactIsPlaneWithPaths` post-merge * Need to disable the sketch-on-face case now that artifactGraph is in Rust. Will active in a future PR (cc @jtran) * Re-run CI after snapshots * Update FeatureTreePane to not use `useCommandsContext`, missed during merge * Fix merge issue, import location change on edited file * fix click test step, which I believe is waiting for context scripts to load * Convert toolbarFixture.exeIndicator to getter We need to convert all these selectors on fixtures to getters, because they can go stale if called on the fixture constructor. * Missed a dumb little thing in toolbarFixture.ts * Fix goof with merge * fmt * Another dumb missed thing during merge I gotta get used to the LazyGit merge tool I'm not good at it yet * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Conver sceneFixture's exeIndicator to a getter Locators on fixtures will be frozen from the time of the fixture's initialization, I'm increasingly convinced * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Post-kwargs E2E test cleanup * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Pierre Jacquier <pierrejacquier39@gmail.com> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2025-02-05 19:33:40 -05:00
if (!('codeRef' in a && a.codeRef)) return false
Selections Refactor (#4381) * selection stuff * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest) * trigger CI * fix bugs * some edge cut stuff * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * fix sketch mode issues * fix more tests, selection in sketch related * more test fixing * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger ci * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger ci * more sketch mode selection fixes * fix unit tests * rename function * remove .only * migrate a more selections types * migrate a more selections types * migrate a more selections types * lint * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * fix bad pathToNode issue * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * fix sketch on face * migrate a more selections types * migrate a more selections types * fix code selection of fillets * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * fix bad path to node, looks like a race * migrate a more selections types * migrate a more selections types * fix cmd bar selections * fix cmd bar selections * fix display issues * migrate a more selections types * Revert "migrate a more selections types" This reverts commit 0d0e453bbbb82d10472b6c9e753c53c8b7e3609d. * migrate a more selections types * clean up1 * clean up 2 * fix types after main merge * review tweaks * fix wall selection bug * Update src/lang/std/engineConnection.ts Co-authored-by: Jonathan Tran <jonnytran@gmail.com> * add franks TODO comment * fix type after main merge, plus a touch of clean up --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2024-11-21 15:04:30 +11:00
return isOverlap(a.codeRef.range, segmentRange)
})
return {
codeRef: codeRefFromRange(segmentRange, ast),
artifact: maybeArtifact ? maybeArtifact[1] : undefined,
}
}),
otherSelections: [],
}
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
// apply edge treatment to selection
const result = modifyAstWithEdgeTreatmentAndTag(
ast,
selection,
parameters,
dependencies
)
if (err(result)) {
return result
}
const { modifiedAst } = result
const newCode = recast(modifiedAst)
expect(newCode).toContain(expectedCode)
}
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
const runDeleteEdgeTreatmentTest = async (
code: string,
edgeTreatmentSnippet: string,
expectedCode: string
) => {
// parse ast
const ast = assertParse(code)
// update artifact graph
await kclManager.executeAst({ ast })
const artifactGraph = engineCommandManager.artifactGraph
// define snippet range
const edgeTreatmentRange = topLevelRange(
code.indexOf(edgeTreatmentSnippet),
code.indexOf(edgeTreatmentSnippet) + edgeTreatmentSnippet.length
)
// find artifact
const maybeArtifact = [...artifactGraph].find(([, artifact]) => {
if (!('codeRef' in artifact)) return false
return isOverlap(artifact.codeRef.range, edgeTreatmentRange)
})
// build selection
const selection: Selection = {
codeRef: codeRefFromRange(edgeTreatmentRange, ast),
artifact: maybeArtifact ? maybeArtifact[1] : undefined,
}
// delete edge treatment
const result = await deleteEdgeTreatment(ast, selection)
if (err(result)) {
return result
}
// recast and check
const newCode = recast(result)
expect(newCode).toContain(expectedCode)
}
const createFilletParameters = (radiusValue: number): FilletParameters => ({
type: EdgeTreatmentType.Fillet,
radius: {
valueAst: createLiteral(radiusValue),
valueText: radiusValue.toString(),
valueCalculated: radiusValue.toString(),
},
})
const createChamferParameters = (lengthValue: number): ChamferParameters => ({
type: EdgeTreatmentType.Chamfer,
length: {
valueAst: createLiteral(lengthValue),
valueText: lengthValue.toString(),
valueCalculated: lengthValue.toString(),
},
})
// Iterate tests over all edge treatment types
Object.values(EdgeTreatmentType).forEach(
(edgeTreatmentType: EdgeTreatmentType) => {
// create parameters based on the edge treatment type
let parameterName: string
let parameters: EdgeTreatmentParameters
if (edgeTreatmentType === EdgeTreatmentType.Fillet) {
parameterName = 'radius'
parameters = createFilletParameters(3)
} else if (edgeTreatmentType === EdgeTreatmentType.Chamfer) {
parameterName = 'length'
parameters = createChamferParameters(3)
} else {
// Handle future edge treatments
return new Error(`Unsupported edge treatment type: ${edgeTreatmentType}`)
}
// run tests
describe(`Testing modifyAstCloneWithEdgeTreatmentAndTag with ${edgeTreatmentType}s`, () => {
it(`should add a ${edgeTreatmentType} to a specific segment`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)`
const segmentSnippets = ['line(end = [0, -20])']
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20], tag = $seg01)
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg01])`
await runModifyAstCloneWithEdgeTreatmentAndTag(
code,
segmentSnippets,
parameters,
expectedCode
)
})
it(`should add a ${edgeTreatmentType} to the sketch pipe`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> extrude(length = -15)`
const segmentSnippets = ['line(end = [0, -20])']
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20], tag = $seg01)
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
|> extrude(length = -15)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg01])`
await runModifyAstCloneWithEdgeTreatmentAndTag(
code,
segmentSnippets,
parameters,
expectedCode
)
})
it(`should add a ${edgeTreatmentType} to an already tagged segment`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20], tag = $seg01)
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)`
const segmentSnippets = ['line(end = [0, -20], tag = $seg01)']
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20], tag = $seg01)
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg01])`
await runModifyAstCloneWithEdgeTreatmentAndTag(
code,
segmentSnippets,
parameters,
expectedCode
)
})
it(`should add a ${edgeTreatmentType} with existing tag on other segment`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)`
const segmentSnippets = ['line(end = [-20, 0])']
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg02)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg02])`
await runModifyAstCloneWithEdgeTreatmentAndTag(
code,
segmentSnippets,
parameters,
expectedCode
)
})
it(`should add a ${edgeTreatmentType} with existing fillet on other segment`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> fillet( radius = 5, tags = [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
const segmentSnippets = ['line(end = [-20, 0])']
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg02)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> fillet(radius = 5, tags = [seg01])
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg02])`
await runModifyAstCloneWithEdgeTreatmentAndTag(
code,
segmentSnippets,
parameters,
expectedCode
)
})
it(`should add a ${edgeTreatmentType} with existing chamfer on other segment`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> chamfer(length = 5, tags = [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
const segmentSnippets = ['line(end = [-20, 0])']
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg02)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> chamfer(length = 5, tags = [seg01])
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg02])`
await runModifyAstCloneWithEdgeTreatmentAndTag(
code,
segmentSnippets,
parameters,
expectedCode
)
})
it(`should add a ${edgeTreatmentType} to two segments of a single extrusion`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)`
const segmentSnippets = ['line(end = [20, 0])', 'line(end = [-20, 0])']
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg02)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg01, seg02])`
await runModifyAstCloneWithEdgeTreatmentAndTag(
code,
segmentSnippets,
parameters,
expectedCode
)
})
it(`should add ${edgeTreatmentType}s to two bodies`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
sketch002 = startSketchOn('XY')
|> startProfileAt([30, 10], %)
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 = [15, 0])
|> line(end = [0, -15])
|> line(end = [-15, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude002 = extrude(sketch002, length = -25)` // <--- body 2
const segmentSnippets = [
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 = [20, 0])',
'line(end = [-20, 0])',
'line(end = [0, -15])',
]
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg02)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg01, seg02])
sketch002 = startSketchOn('XY')
|> startProfileAt([30, 10], %)
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 = [15, 0])
|> line(end = [0, -15], tag = $seg03)
|> line(end = [-15, 0])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude002 = extrude(sketch002, length = -25)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg03])` // <-- able to add a new one
await runModifyAstCloneWithEdgeTreatmentAndTag(
code,
segmentSnippets,
parameters,
expectedCode
)
})
})
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
describe(`Testing deleteEdgeTreatment with ${edgeTreatmentType}s`, () => {
// simple cases
it(`should delete a piped ${edgeTreatmentType} from a single segment`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg01)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg01])`
const edgeTreatmentSnippet = `${edgeTreatmentType}(${parameterName} = 3, tags = [seg01])`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg01)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
await runDeleteEdgeTreatmentTest(
code,
edgeTreatmentSnippet,
expectedCode
)
})
it(`should delete a non-piped ${edgeTreatmentType} from a single segment`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg01)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
fillet001 = ${edgeTreatmentType}(extrude001, ${parameterName} = 3, tags = [seg01])`
const edgeTreatmentSnippet = `fillet001 = ${edgeTreatmentType}(extrude001, ${parameterName} = 3, tags = [seg01])`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg01)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
await runDeleteEdgeTreatmentTest(
code,
edgeTreatmentSnippet,
expectedCode
)
})
// getOppositeEdge and getNextAdjacentEdge cases
it(`should delete a piped ${edgeTreatmentType} tagged with getOppositeEdge`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg01)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
fillet001 = ${edgeTreatmentType}(extrude001, ${parameterName} = 3, tags = [getOppositeEdge(seg01)])`
const edgeTreatmentSnippet = `fillet001 = ${edgeTreatmentType}(extrude001, ${parameterName} = 3, tags = [getOppositeEdge(seg01)])`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg01)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
await runDeleteEdgeTreatmentTest(
code,
edgeTreatmentSnippet,
expectedCode
)
})
it(`should delete a non-piped ${edgeTreatmentType} tagged with getNextAdjacentEdge`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg01)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
fillet001 = ${edgeTreatmentType}(extrude001, ${parameterName} = 3, tags = [getNextAdjacentEdge(seg01)])`
const edgeTreatmentSnippet = `fillet001 = ${edgeTreatmentType}(extrude001, ${parameterName} = 3, tags = [getNextAdjacentEdge(seg01)])`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0])
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg01)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
await runDeleteEdgeTreatmentTest(
code,
edgeTreatmentSnippet,
expectedCode
)
})
// cases with several edge treatments
it(`should delete a piped ${edgeTreatmentType} from a body with multiple treatments`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg02)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg01])
|> fillet(radius = 5, tags = [getOppositeEdge(seg02)])
fillet001 = ${edgeTreatmentType}(extrude001, ${parameterName} = 6, tags = [seg02])
chamfer001 = chamfer(extrude001, length = 5, tags = [getOppositeEdge(seg01)])`
const edgeTreatmentSnippet = `${edgeTreatmentType}(${parameterName} = 3, tags = [seg01])`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg02)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> fillet(radius = 5, tags = [getOppositeEdge(seg02)])
fillet001 = ${edgeTreatmentType}(extrude001, ${parameterName} = 6, tags = [seg02])
chamfer001 = chamfer(extrude001, length = 5, tags = [getOppositeEdge(seg01)])`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
await runDeleteEdgeTreatmentTest(
code,
edgeTreatmentSnippet,
expectedCode
)
})
it(`should delete a non-piped ${edgeTreatmentType} from a body with multiple treatments`, async () => {
const code = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg02)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg01])
|> fillet( radius = 5, tags = [getOppositeEdge(seg02)] )
fillet001 = ${edgeTreatmentType}(extrude001, ${parameterName} = 6, tags = [seg02])
chamfer001 = chamfer(extrude001, length = 5, tags = [getOppositeEdge(seg01)])`
const edgeTreatmentSnippet = `fillet001 = ${edgeTreatmentType}(extrude001, ${parameterName} = 6, tags = [seg02])`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
const expectedCode = `sketch001 = startSketchOn('XY')
|> startProfileAt([-10, 10], %)
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 = [20, 0], tag = $seg01)
|> line(end = [0, -20])
|> line(end = [-20, 0], tag = $seg02)
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -15)
|> ${edgeTreatmentType}(${parameterName} = 3, tags = [seg01])
|> fillet(radius = 5, tags = [getOppositeEdge(seg02)])
chamfer001 = chamfer(extrude001, length = 5, tags = [getOppositeEdge(seg01)])`
Add Point-and-Click Deletion for Fillets and Chamfers (#5098) * ast mod * point and click test * tsc * test test * unit test edit * topLevelRange * disable unit test * remove bad imports * fix typo * Fix cyclic dependency hell with getNodePathFromSourceRange * tsc * fix ImportStatement * fix isValueZero * pre-emptively ==> preemptively * yarn fmt-check * reenable the unit test * fmt * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * add test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * several treatments * consolidate * typos * fix imports, consolidate * consolidate import * fix imports * add tests * stress test CI * fix test * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * fix tests * clean test for fillets * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * Trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: namespace-profile-ubuntu-8-cores) * test chamfers * comments * simplify main tests * typo * typo2 * remove import * clean up comments --------- Co-authored-by: 49lf <ircsurfer33@gmail.com> Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
2025-02-03 18:11:26 +01:00
await runDeleteEdgeTreatmentTest(
code,
edgeTreatmentSnippet,
expectedCode
)
})
})
}
)
describe('Testing isTagUsedInEdgeTreatment', () => {
const code = `sketch001 = startSketchOn('XZ')
|> startProfileAt([7.72, 4.13], %)
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 = [7.11, 3.48], tag = $seg01)
|> line(end = [-3.29, -13.85])
|> line(end = [-6.37, 3.88], tag = $seg02)
|> close()
extrude001 = extrude(sketch001, length = -5)
|> fillet(
radius = 1.11,
tags = [
Tag enhancements (#3143) * start Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * more Signed-off-by: Jess Frazelle <github@jessfraz.com> * enhancements Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * get plane data Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * cleanup Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * more Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * fmt Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * add lint rule Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> --------- Signed-off-by: Jess Frazelle <github@jessfraz.com>
2024-07-27 22:56:46 -07:00
getOppositeEdge(seg01),
seg01,
Tag enhancements (#3143) * start Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * more Signed-off-by: Jess Frazelle <github@jessfraz.com> * enhancements Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * get plane data Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * cleanup Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * more Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * fmt Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * updates Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * add lint rule Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> * fixes Signed-off-by: Jess Frazelle <github@jessfraz.com> --------- Signed-off-by: Jess Frazelle <github@jessfraz.com>
2024-07-27 22:56:46 -07:00
getPreviousAdjacentEdge(seg02)
]
)
`
it('should correctly identify getOppositeEdge and baseEdge edges', () => {
const ast = assertParse(code)
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 lineOfInterest = `line(end = [7.11, 3.48], tag = $seg01)`
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
const range = topLevelRange(
code.indexOf(lineOfInterest),
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
code.indexOf(lineOfInterest) + lineOfInterest.length
)
const pathToNode = getNodePathFromSourceRange(ast, range)
if (err(pathToNode)) return
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 callExp = getNodeFromPath<CallExpression | CallExpressionKw>(
ast,
pathToNode,
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
['CallExpression', 'CallExpressionKw']
)
if (err(callExp)) return
const edges = isTagUsedInEdgeTreatment({ ast, callExp: callExp.node })
expect(edges).toEqual(['getOppositeEdge', 'baseEdge'])
})
it('should correctly identify getPreviousAdjacentEdge edges', () => {
const ast = assertParse(code)
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 lineOfInterest = `line(end = [-6.37, 3.88], tag = $seg02)`
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
const range = topLevelRange(
code.indexOf(lineOfInterest),
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
code.indexOf(lineOfInterest) + lineOfInterest.length
)
const pathToNode = getNodePathFromSourceRange(ast, range)
if (err(pathToNode)) return
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 callExp = getNodeFromPath<CallExpression | CallExpressionKw>(
ast,
pathToNode,
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
['CallExpression', 'CallExpressionKw']
)
if (err(callExp)) return
const edges = isTagUsedInEdgeTreatment({ ast, callExp: callExp.node })
expect(edges).toEqual(['getPreviousAdjacentEdge'])
})
it('should correctly identify no edges', () => {
const ast = assertParse(code)
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 lineOfInterest = `line(end = [-3.29, -13.85])`
const start = code.indexOf(lineOfInterest)
expect(start).toBeGreaterThan(-1)
const range = topLevelRange(start, start + lineOfInterest.length)
const pathToNode = getNodePathFromSourceRange(ast, range)
if (err(pathToNode)) return
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 callExp = getNodeFromPath<CallExpressionKw>(
ast,
pathToNode,
'CallExpression'
)
if (err(callExp)) return
const edges = isTagUsedInEdgeTreatment({ ast, callExp: callExp.node })
expect(edges).toEqual([])
})
})
describe('Testing button states', () => {
const runButtonStateTest = async (
code: string,
segmentSnippet: string,
expectedState: boolean
) => {
const ast = assertParse(code)
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 start = code.indexOf(segmentSnippet)
expect(start).toBeGreaterThan(-1)
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
const range = segmentSnippet
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
? topLevelRange(start, start + segmentSnippet.length)
Rust artifact graph (#5068) * Start porting artifact graph creation to Rust * Add most of artifact graph creation * Add handling loft command from recent PR * Refactor artifact merge code so that it errors when a new artifact type is added * Add sweep subtype * Finish implementation of build artifact graph * Fix wasm.ts to use new combined generated ts-rs file * Fix Rust lints * Fix lints * Fix up replacement code * Add artifact graph to WASM outcome * Add artifact graph to simulation test output * Add new artifact graph output snapshots * Fix wall field and reduce unreachable code * Change field order for subtype * Change subtype to be determined from the request, like the TS * Fix plane sweep_id * Condense code * Change ID types to be properly optional * Change to favor the new ID, the same as TS * Fix to make error impossible * Rename artifact type tag values to match TS * Fix name of field on Cap * Update outputs * Change to use Rust source range * Update output snapshots * Add conversion to mermaid mind map and add to snapshot tests * Add new mermaid mind map output * Add flowchart * Remove raw artifact graph from tests * Remove JSON artifact graph output * Update output file with header * Update output after adding flowchart * Fix flowchart to not have duplicate edges, one in each direction * Fix not not output duplicate edges in flowcharts * Change flowchart edge style to be more obvious when a direction is missing * Update output after deduplication of edges * Fix not not skip sketch-on-face artifacts * Add docs * Fix edge iteration order to be stable * Update output after fixing order * Port TS artifactGraph.test.ts tests to simulation tests * Add grouping segments and solid2ds with their path * Update output flowcharts since grouping paths * Remove TS artifactGraph tests * Remove unused d3 dependencies * Fix to track loft ID on paths * Add command ID to error messages * Move artifact graph test code to a separate file since it's a large file * Reduce function visibility * Remove TS artifact graph code * Fix spelling error with serde * Add TODO for edge cut consumed ID * Add comment about mermaid edge rank * Fix mermaid flowchart edge cuts to appear as children of their edges * Update output since fixing flowchart order * Fix to always build the artifact graph even when there's a KCL error * Add artifact graph to error output * Change optional ID merge to match TS * Remove redundant SourceRange definition * Remove Rust-flavored default source range function * Add helper for source range creation * Update doc comment for the website * Update docs after doc comment change * Fix to save engine responses in execution cache * Remove unused import * Fix to not call WASM function before beforeAll callback is run * Remove more unused imports
2025-01-17 14:34:36 -05:00
: topLevelRange(ast.end, ast.end) // empty line in the end of the code
const selectionRanges: Selections = {
Selections Refactor (#4381) * selection stuff * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest) * trigger CI * fix bugs * some edge cut stuff * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * fix sketch mode issues * fix more tests, selection in sketch related * more test fixing * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger ci * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger ci * more sketch mode selection fixes * fix unit tests * rename function * remove .only * migrate a more selections types * migrate a more selections types * migrate a more selections types * lint * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * fix bad pathToNode issue * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * fix sketch on face * migrate a more selections types * migrate a more selections types * fix code selection of fillets * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * fix bad path to node, looks like a race * migrate a more selections types * migrate a more selections types * fix cmd bar selections * fix cmd bar selections * fix display issues * migrate a more selections types * Revert "migrate a more selections types" This reverts commit 0d0e453bbbb82d10472b6c9e753c53c8b7e3609d. * migrate a more selections types * clean up1 * clean up 2 * fix types after main merge * review tweaks * fix wall selection bug * Update src/lang/std/engineConnection.ts Co-authored-by: Jonathan Tran <jonnytran@gmail.com> * add franks TODO comment * fix type after main merge, plus a touch of clean up --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2024-11-21 15:04:30 +11:00
graphSelections: [
{
Selections Refactor (#4381) * selection stuff * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest) * trigger CI * fix bugs * some edge cut stuff * A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu-latest-8-cores) * trigger CI * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * fix sketch mode issues * fix more tests, selection in sketch related * more test fixing * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger ci * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * Trigger ci * more sketch mode selection fixes * fix unit tests * rename function * remove .only * migrate a more selections types * migrate a more selections types * migrate a more selections types * lint * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * fix bad pathToNode issue * A snapshot a day keeps the bugs away! 📷🐛 (OS: windows-latest-8-cores) * fix sketch on face * migrate a more selections types * migrate a more selections types * fix code selection of fillets * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * migrate a more selections types * fix bad path to node, looks like a race * migrate a more selections types * migrate a more selections types * fix cmd bar selections * fix cmd bar selections * fix display issues * migrate a more selections types * Revert "migrate a more selections types" This reverts commit 0d0e453bbbb82d10472b6c9e753c53c8b7e3609d. * migrate a more selections types * clean up1 * clean up 2 * fix types after main merge * review tweaks * fix wall selection bug * Update src/lang/std/engineConnection.ts Co-authored-by: Jonathan Tran <jonnytran@gmail.com> * add franks TODO comment * fix type after main merge, plus a touch of clean up --------- Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com> Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
2024-11-21 15:04:30 +11:00
codeRef: codeRefFromRange(range, ast),
},
],
otherSelections: [],
}
// state
const buttonState = hasValidEdgeTreatmentSelection({
ast,
selectionRanges,
code,
})
expect(buttonState).toEqual(expectedState)
}
const codeWithBody: string = `
sketch001 = startSketchOn('XY')
|> startProfileAt([-20, -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
|> line(end = [0, 10])
|> line(end = [10, 0])
|> line(end = [0, -10])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
extrude001 = extrude(sketch001, length = -10)
`
const codeWithoutBodies: string = `
sketch001 = startSketchOn('XY')
|> startProfileAt([-20, -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
|> line(end = [0, 10])
|> line(end = [10, 0])
|> line(end = [0, -10])
|> line(endAbsolute = [profileStartX(%), profileStartY(%)])
|> close()
`
// body is missing
it('should return false when body is missing and nothing is selected', async () => {
await runButtonStateTest(codeWithoutBodies, '', false)
})
it('should return false when body is missing and segment is selected', async () => {
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
await runButtonStateTest(codeWithoutBodies, `line(end = [10, 0])`, false)
})
// body exists
it('should return true when body exists and nothing is selected', async () => {
await runButtonStateTest(codeWithBody, '', true)
})
it('should return true when body exists and segment is selected', async () => {
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
await runButtonStateTest(codeWithBody, `line(end = [10, 0])`, true)
})
it('should return false when body exists and not a segment is selected', async () => {
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
await runButtonStateTest(codeWithBody, `close()`, false)
})
})