On `main`, the new test program in `tests/executor` causes a stack overflow. Running a flamegraph via `sudo cargo flamegraph --test executor -- serial_test_mike_stress_lines` shows that the problem: `fn execute_pipe_body` is very recursive.
When there's a long pipe, `execute_pipe_body` executes the next child expression, which is always a CallExpression. So `execute_pipe_body` calls `CallExpression::execute`. Which then calls `execute_pipe_body` again.
Fix is simple: `execute_pipe_body` now iterates over pipe subexpressions instead of recursing. This fixes the stack overflow, and is much faster too.
Closes https://github.com/KittyCAD/modeling-app/issues/1891.
* Make tag last optional param
* Update all test assertions with correct tag format
* Format ts
* Some progress on tests and code mods
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* More sketch fixes
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Only 1 test left
* Clean up console.log
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* Fix last ts test
* Clean up fmt
* Fix clippy too
* Update docs and fix small oversight on angled lines
* Fix more rust tests
* Make typescript happy
* Fmt
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Co-authored-by: Jess Frazelle <jessfraz@users.noreply.github.com>
* updates
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>
* make tsc happ
* better error msg
* fix control point issue
* basic code gen working for tangentialArc
* partical fix for move with arcs
* tangential arc move
* fix
* make eslint rules less annoying
* inital refactor of some xstate stuff
* more old tangential arc clean up stuff
* more tweaks
* add testing
* tweak xstate inspect
* temp remove test
* update formating for less conflicts
* fix state machine layout after merge
* shrug, something weird with xstate typegen
* renaming some xstate events
* tweak numbers to make CI playwright happ
* CI hacks
* more CI hacks
* more CI hacks
* new hack strategy
* run tests agian
* make cmd bar less flaky
* ci hacks
* CI hacks
* CI hacks
* CI hacks
* clean up
* fix
* still have constraint stuff to deal with
* progress on move rules
* update source ranges after no execute code-mod
* typo
* mvp working
* hide show sketch overlay
* match scaling
* update arrow head style
* animate line tool
* bypass xstate for animations, much smoother
* add new segment working with refactor needed for setup paper sketch
* refactor setup paper sketch
* tangantialArcTo drag animations working
* tangential arc polish
* cargo fmt
* clippy
* more clippy
* mock canvas
* last of clippy?
* typo
* more clippy stuff
* move util function so they are shareable with typescript
* migrate a bunch to rust and only rust
* add arc center point for draft tangential ac
* clippy tweak
* delete uneeded test
* Rough start to scaling arrow heads.
The tangent arrow heads are basically nuked and replaced while the
straight line sections are just rotated and repositioned, this means they
miss out on updating scaling number after a screen size changes.
Needs fixing
* fix bug with tool tips
* fix draft line start position
Having drag the end of teh path before selecting a tooltip would result in the draft line starting where the path used to end, stale data
* some progress with pan maybe
* fmt
* inital camera sync working
For perspective camera at least
* change three.js to use z-up
* add grid
* orthographic camera working with polish items TODO
* fix zoom level when swapping camera
* fix up camera/orbit changing on cam change (pan wasn't being respected)
* tidy up
* use orbit target instead of assuming scene center
* dynamic fov working
* animate orthographic to perspective and reverse
* fix import
* temp fix for batch commands
* initial client side scene sketch working
* remove hover log
* FOV adjust fix
* fix comment
* tear down sketch and small tweaks
* some progress with camera tweening
* combine dollyZoom engine commands
see
https://github.com/KittyCAD/modeling-api/compare/kurt-perspective-settings?expand=1
and
https://github.com/KittyCAD/engine/compare/kurt-perspective-settings?expand=1
* make tests happy (mocks)
* fix tween to vertical/camera-up bug
* tween to each axis with hacky solutions in there
* fix startSketchOn planes
* tidy startSketchOn
* tweening okay for now I think
* get sketching on default planes working
* allow editing on all default planes
* clean up enter and exit sketch logic
* tidy
* tidy
* remove more default plane stuff
* start of draft line
* remove some annoying parts of the paper.js implementation
* fix drag than equip line bug
* comment
* don't animate on skech tear down since it's used for draft line
* remove more default plane shit
* style draft line
* refine dashed line
* draft line set up and tear down mostly happy
* add on click logic ready for draft lines
* sketch mode with drag and draft mode working solidly now, straight segments only
* default planes match colors, hover and select still TODO
* hover and click logic working for default planes
Now just need the code mode to fire to 'startSketchOn(...)'
* select default planes
* remove some logs
* fix update infinite loop
* start of orbitControls port to Franks control guards
* hiding scenes at different times
* scene hide on camera move should be respected by scroll zoom
* basic hover working
* Hook up user camera settings to ClientSideScene (#1334)
* Refactor to not import utilities from Router.tsx
* Stop tracking changes or formatting *.typegen.ts
* Hook up cameraControls to ClientSideScene
* Remove camera controls toggle from temp debug panel
---------
Co-authored-by: Kurt Hutten Irev-Dev <k.hutten@protonmail.ch>
* add select segment moves cursor
* highlight segments yellow on hover
* cursor ranges effect 2d line colors
* fix constrainst i.e. make sure the sketch is rejiged
* selecting nothing should remove selections
* remove hardcoded strings
* update get_tangential_arc_to_info rust util
* initial drawing of tangential arcs in client scene
* fix tangentialArc arrow head direction
* correct userData types for tangential arcs
* get tangential arc updates working
Doesn't include draging the head of the tangential arc itself yet
* spot of clean up
* make selections work with tangential arcs
* get draft tangential segment animated
* fix initial click weirdness for adding new tangential line
* couple tweaks
* add grace pixels /threshold to raycast
* redo arc dashes so that they spawn from the ccenter of the arc
* fix multi drag bug
* fmt
* add temp solution for close
* add default axis hover colors, still needs select logic
* selection of axis works, just with out selection color
* get axis selection colors working
* fix outdate source ranges after drag problem
* update moreNodePathFromSourceRange
* fix ts-rs issue/workaround
* fix default plane weirdness
* fix tangential arc rounding issue
* review clean up part 1
* review clean up part 2
Big state-diagram cull
* clippy
* typo
* clippy
* fix xstate types with typegen
* fix types
* clippy
* catch error
* fix test import issue
Not sure exactly what was happening but guessing circular import that vite didn't like
* add axis/plane info to sketch group tests
* case changes because of rs-ts bug, can probably revert this later
* start of playwright test fixes
* reduce geo complexity for straight segments
* fix cam adjust tests
* Revert "Clean up vite build warnings (#1332)"
This reverts commit c1f661ab52.
* selection e2e test fixed<
* remove camToggle to allow playwright tests to pass
* remove drag test
too brittle and needs to be redone from the ground up anyway
* trigger CI
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* fix last test
* clean up part 3
* clean up part 4
* clean up part 5
* clean up sketch enter exit logic
* fix engine side selections
* default plane should not be selected form 'onDragEnd'
i.e. rotating the camera should not mean the user acidently selects a plane
* clean up state diagram around animating to sketch mode
Embracing that the animation is async and puting the interdiate steps in the state diagram clean up some logic and solved some bugs at the same time
* add test for multiple sketches
* typo
* make highlight more robust
* type tweak
* scale segmenst with distance from camera so they have a consistent pixel size/ screenspace size
* Jess's advice
* tsc and fmt
* clean up part 6
remove integer from xstate names
* clean up part 7
* integrate sequency in to camera moves
* fix tests
* update snapshot e2e
* small snapshot change
* A snapshot a day keeps the bugs away! 📷🐛 (OS: ubuntu)
* trigger ci
* Fix HomeLoaderData types
* update std stuff
* update kittycad rs client lib
---------
Signed-off-by: Jess Frazelle <github@jessfraz.com>
Co-authored-by: Jess Frazelle <github@jessfraz.com>
Co-authored-by: Frank Noirot <frank@kittycad.io>
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
# Problem
This is my proposal for fixing #1107 . I've only done it for one stdlib function, `tangentialArcTo` -- if y'all like it, I'll apply this idea to the rest of the stdlib.
Previously, if users want to put a tag on the arc, the function's parameters change type.
```
// Tag missing: first param is array
tangentialArcTo([x, y], %)
// Tag present: first param is object
tangentialArcTo({to: [x, y], tag: "myTag"}, %)
```
# Solution
My proposal in #1006 is that KCL should have optional values. This means we can change the stdlib `tangentialArcTo` function to use them. In this PR, the calls are now like
```
// Tag missing: first param is array
tangentialArcTo([x, y], %)
// Tag present: first param is array still, but we now pass a tag at the end.
tangentialArcTo([x, y], %, "myTag")
```
This adds an "option" type to KCL typesystem, but it's not really revealed to users (no KCL types are revealed to users right now, they write untyped code and only interact with types when they get type errors upon executing programs). Also adds a None type, which is the default case of the Optional enum.
* AST function nodes no longer have stdlib function members
IMO the AST should not need an actual pointer to a stdlib function -- that is a completely separate concern from the AST.
Instead, the AST nodes can just store function names, and the executor will have a stdlib which it can look up those names in.
* Fix tests
* Update snapshot tests
* New parser built with Winnow
This new parser uses [winnow](docs.rs/winnow) to replace the handwritten recursive parser.
## Differences
I think the Winnow parser is more readable than handwritten one, due to reusing standard combinators. If you have a parsre like `p` or `q` you can combine them with standard functions like `repeat(0..4, p)`, `opt(p)`, `alt((p, q))` and `separated1(p, ", ")`. This IMO makes it more readable once you know what those standard functions do.
It's also more accurate now -- e.g. the parser no longer swallows whitespace between comments, or inserts it where there was none before. It no longer changes // comments to /* comments depending on the surrounding whitespace.
Primary form of testing was running the same KCL program through both the old and new parsers and asserting that both parsers produce the same AST. See the test `parser::parser_impl::tests::check_parsers_work_the_same`. But occasionally the new and old parsers disagree. This is either:
- Innocuous (e.g. disagreeing on whether a comment starts at the preceding whitespace or at the //)
- Helpful (e.g. new parser recognizes comments more accurately, preserving the difference between // and /* comments)
- Acceptably bad (e.g. new parser sometimes outputs worse error messages, TODO in #784)
so those KCL programs have their own unit tests in `parser_impl.rs` demonstrating the behaviour.
If you'd like to review this PR, it's arguably more important to review changes to the existing unit tests rather than the new parser itself. Because changes to the unit tests show where my parser changes behaviour -- usually for the better, occasionally for the worse (e.g. a worse error message than before). I think overall the improvements are worth it that I'd like to merge it without spending another week fixing it up -- we can fix the error messages in a follow-up PR.
## Performance
| Benchmark | Old parser (this branch) | New parser (this branch) | Speedup |
| ------------- | ------------- | ------------- | ------------- |
| Pipes on pipes | 922 ms | 42 ms | 21x |
| Kitt SVG | 148 ms | 7 ms | 21x |
There's definitely still room to improve performance:
- https://github.com/KittyCAD/modeling-app/issues/839
- https://github.com/KittyCAD/modeling-app/issues/840
## Winnow
Y'all know I love [Nom](docs.rs/nom) and I've blogged about it a lot. But I'm very happy using Winnow, a fork. It's got some really nice usability improvements. While writing this PR I found some bugs or unclear docs in Winnow:
- https://github.com/winnow-rs/winnow/issues/339
- https://github.com/winnow-rs/winnow/issues/341
- https://github.com/winnow-rs/winnow/issues/342
- https://github.com/winnow-rs/winnow/issues/344
The maintainer was quick to close them and release new versions within a few hours, so I feel very confident building the parser on this library. It's a clear improvement over Nom and it's used in toml-edit (and therefore within Cargo) and Gitoxide, so it's becoming a staple of the Rust ecosystem, which adds confidence.
Closes#716Closes#815Closes#599
* Remove duplicated tests
These tests already were copied to tokeniser2.rs, so removing them doesn't affect code coverage.
* Move tokeniser to its own module
Now there's a module for tokens, and the tokenizer/lexer implementation is private within the token module.
