Fix lack of error on default export of nonexistant name #40094
Merged
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Fixes #39694
This was actually more interesting to fix than you may think, because it didn't repro in the test harness. Turn out that's because the test harness calls
emit
, and thengetPreEmitDiagnostics
, which is different from the LS or command line compiler - namely, it captures an error added by theEmitResolver
(which shouldn't be adding new errors in the first place)! So once I realized this, I added some new test infrastructure to figure out when this happens and report it (by getting the diagnostics twice - once beforeemit
and once after). That then exposed a crash that wasn't captured by our test harness but would have happened on the command line (accessing a parent pointer in a place where it may not have been bound), which I fixed, and exposed two existing tests where similarEmitResolver
-added-error behavior is occurring. In one case, the order we checked controlled which declarations we actually recorded duplicate identifier errors on, because we were only scanning for them on one "side" of the conflicting declarations. In the other, we were explicitly only looking for flag conflicts on the "first" declaration bound to a symbol, but that meant that if checking started in a file other than the one that declaration was in, we'd fail to capture any errors for the conflict.