When declaring a function, its first parameter is allowed to be prefixed with `@`. This means that when users call this function, they don't have to label this argument.
Only the first parameter is allowed this prefix, no others.
Part of https://github.com/KittyCAD/modeling-app/issues/4600
* fix: make variable declaration errors Cut instead of Backtrace
* fix: clippy, move comma to empty case and add test
* fix: add missing TokenType case
* fix: incorrect fn args after merge
* fix: clippy lint
* fix: update error message being looked for in e2e test
---------
Co-authored-by: Tom Pridham <pridham.tom@gmail.com>
Co-authored-by: Jonathan Tran <jonnytran@gmail.com>
* Parse more import syntax
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* Remove unnecessary Vec from VariableDeclaration
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* Parse export import
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* Factor out an execution module
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* imports: constants, globs, export import
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* test fixups
Signed-off-by: Nick Cameron <nrc@ncameron.org>
---------
Signed-off-by: Nick Cameron <nrc@ncameron.org>
Pure refactor, should not change any behaviour.
Previously, optional parameters in KCL function calls always set the parameter to KclNone.
As of this PR, they can be set to KCL literals in addition to KCL none. However the parser does not actually ever use this (that'll be in a follow-up PR).
Also adds a `labeled: bool` to all parameters, which is always true. But it lays the groundwork for the unlabeled first parameter in a follow-up PR.
* Send multiple errors and warnings to the frontend and LSP
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* Refactor the parser to use CompilationError for parsing errors rather than KclError
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* Refactoring: move CompilationError, etc.
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* Integrate compilation errors with the frontend and CodeMirror
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* Fix tests
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* Review comments
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* Fix module id/source range stuff
Signed-off-by: Nick Cameron <nrc@ncameron.org>
* More test fixups
Signed-off-by: Nick Cameron <nrc@ncameron.org>
---------
Signed-off-by: Nick Cameron <nrc@ncameron.org>
We don't have any of these, and I don't think it's
worth the complexity. The goal was to let us write
KCL stdlib functions in KCL not Rust. But who cares
really. We can always put this back if we need it.
* Fix broken test from previous PR
* Look at this (photo)Graph *in the voice of Nickelback*
* void
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
Part of https://github.com/KittyCAD/modeling-app/issues/4600
Adds support for keyword arguments to the stdlib, and calling stdlib functions with keyword arguments.
So far, I've changed one function: `rem`. Previously you would have used `rem(7, 2)` but now it's `rem(7, divisor: 2)`.
This is a proof-of-concept. If it's approved, we will:
1. Support closures with keyword arguments, and calling them
2. Move the rest of the stdlib to use kw arguments
* Fix default planes to be created in deterministic order
* 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)
* Trigger CI
---------
Co-authored-by: github-actions[bot] <github-actions[bot]@users.noreply.github.com>
