Fix signature matching for externals when abstract alias hides function arity#8045
Conversation
|
@cristianoc fix by Codex - could you have a look? |
rescript
@rescript/darwin-arm64
@rescript/darwin-x64
@rescript/linux-arm64
@rescript/linux-x64
@rescript/runtime
@rescript/win32-x64
commit: |
36611cd to
3a78abe
Compare
cristianoc
left a comment
There was a problem hiding this comment.
I think this copied a certain style from existing tests lol.
Opaque test name.
And "regression".
If it is a regression, point out in what compiler version the issue was introduced.
cristianoc
left a comment
There was a problem hiding this comment.
Just asking to clean up: proper name for the test file, and description in the PR title of what is being fixed.
That said, the question is whether it can be fixed early by trying to avoid to create the issue in the first place.
It worked fine in ReScript 11, but broke in ReScript 12.
Can you clarify? I guess the "gpr_<issue_number>" prefix following the conventions from other files is fine?
What would you suggest as the file name / PR title? |
That's some old leftover convention that creates an unnecessary indirection and says nothing about what's tested. |
cristianoc
left a comment
There was a problem hiding this comment.
Ready to go after the suggestions are in.
Still I would ask the agent if this can be prevented earlier instead of fixing just before signature inclusion.
3a78abe to
ceedc3c
Compare
|
@cristianoc Was not successful in preventing earlier instead of fixing just before signature inclusion. I did change the naming though, please have another look. |
Looks great. |
Fixes #8038