Add a test showing how Core can be implicitly poisoned
#4900
Conversation
| // CHECK:STDERR: class Core {} | ||
| // CHECK:STDERR: ^~~~~~~~~~~~ | ||
| // CHECK:STDERR: | ||
| class Core {} |
There was a problem hiding this comment.
| class Core {} | |
| class r#Core {} |
There was a problem hiding this comment.
Note, suggesting this specifically so that making Core a keyword causes a corresponding behavior change in this test. Which I think is probably fine.
FYI @danakj
| @@ -369,6 +369,23 @@ fn F() { | |||
| } | |||
| } | |||
|
|
|||
| // --- fail_implicitly_poison_core.carbon | |||
There was a problem hiding this comment.
Why is this in testdata/function/declaration/...? It doesn't really seem to be testing function declarations.
There was a problem hiding this comment.
You're absolutely right, there a lot of tests here that are definitely not about function declaration and/or not specific to function declaration.
I thought of putting all name poisoning tests together, but I agree this is not a good place.
I think at least some of the tests don't really belong to any specific existing testdata directory.
Should I create a dedicated one? Perhaps "lookup"?
Will probably do this in a separate PR though if that's ok.
There was a problem hiding this comment.
#4909 has testdata/packages/raw_core.carbon. You're testing name poisoning of the package; how about either that file or testdata/pavckages/core_name_poisoning.carbon?
Maybe this is small enough to just take care of in this PR?
There was a problem hiding this comment.
I've moved it to toolchain/check/testdata/packages/no_prelude/core_name_poisoning.carbon.
There was a problem hiding this comment.
Thanks! I see you requested another review from zygoloid, but I think this is small enough to go ahead with a merge.
…`toolchain/check/testdata/packages/no_prelude/core_name_poisoning.carbon`
…istent This is also following carbon-language#4900 (comment), which points that name poisoning tests are not in the correct place.
…istent (#4987) This is also following #4900 (comment), which points that name poisoning tests are not in the correct place. Part of #4622.
Alternatively, if poisoning
Coreis considered a bug, we could avoid poisoning it (see #4903).Part of #4622.