Conversation
Should catch basic `zig fmt` regressions that were previously going uncaught and breaking things
Is this after edea7a4? I did test that commit on windows before merging like this: |
|
Also, I think this change might be able to be much simpler, just by changing Line 50 in 5622044 build.zig to a directory.
|
|
Thanks BTW 👍 |
Yes, after edea7a4. (this is likely #5643) (this is likely caused by some change in edea7a4) I'm still using Windows 7, though, so that could be the difference? EDIT: Don't think this is the problem; I'm expecting the CI to fail as well. EDIT#2: CI is failing, so it's not Windows 7-specific: |
This would still be an incomplete test, since it doesn't actually make changes to the file ( The added cli tests in this PR intentionally create files that will be changed by |
Should help to reduce the number of regressions that have been happening with
zig fmt.This is failing for me locally on Windows with unexpected status 0xC0000056 (
STATUS_DELETE_PENDING), but is working on Linux.zig fmtis misbehaving for me on Windows currently though (AccessDenied or unexpected 0xC0000103 (NOT_A_DIRECTORY)) when invoking it normally, so I think it's catching an actual regression.