Conversation
| @@ -0,0 +1,15 @@ | |||
| [workspace] | |||
There was a problem hiding this comment.
this helps for organizing running tests better. can also be used in the future to optimize compilation on mac, but dont wanna mess with that right now
There was a problem hiding this comment.
we ended up getting improved compilation on mac locally for free with these changes
| mod tests { | ||
| use super::*; | ||
|
|
||
| #[test] |
There was a problem hiding this comment.
a lot of this was vibecoded based on the patterns above just to prevent drift via unit tests
evanmarshall
left a comment
evanmarshall
left a comment
There was a problem hiding this comment.
Mostly questions but I'd like to review it again before merging.
electron-builder.config.js
Outdated
| })) | ||
| nativeBinaries.map(binary => { | ||
| // Map electron-builder arch to Cargo target triple | ||
| const cargoArch = '${arch}' === 'arm64' ? 'aarch64' : 'x86_64' |
There was a problem hiding this comment.
this lets up clean up the logic in build-binaries where we move resources around to different folders.
the removed logic was causing issues with local workspace development with resources not being where they were expected on re-compilation
| .send(command) | ||
| .expect("Failed to send command to processor"); | ||
| } | ||
| for l in stdin.lock().lines().map_while(Result::ok) { |
There was a problem hiding this comment.
clippy recommended this changes, binaries were still functioning the same
2 participants
Extensive unit tests are largely going to be more difficult to implement. These top level ones test non-OS specific calls / business logic. Once we get into OS specific calls, we'd either have to set up a dry-run + mocking conditional to our hot paths to test logic, and/or have our runners be different OS so that they can run the tests for the OS specific logic. Something to tackle in a follow-up ticket