jimingham wrote:

> LGTM. Listing those languages is correct for `GetSupportedLanguagesForTypes` 
> (as opposed to `GetSupportedLanguagesForExpressions`). Even more robust would 
> be to say that `TypeSystemClangSupportsLanguage` should return the union of 
> those two, but in practice that's already the case.
> 
> Normally I'm a big proponent of splitting up PRs if it makes sense and would 
> agree with Michael, but I think this is borderline.
> 
> To Jim's point, hopefully we're already doing the right thing for expressions 
> and checking `GetSupportedLanguagesForExpressions`.

No, that's where the real lie is.  Rust says "I'm C++, look no further" so that:

(lldb) expr -l rust -- 1 + 1
(int) $0 = 2

That only sort of works, since that really is just using the C++ parser to 
parse rust expressions, which only works sometimes.  We should probably not 
error out in this case, since some expressions do work.  But it would be really 
nice to warn people that their rust expressions are getting parsed by a C++ 
parser which doesn't actually know anything about Rust.


https://github.com/llvm/llvm-project/pull/172567
_______________________________________________
lldb-commits mailing list
[email protected]
https://lists.llvm.org/cgi-bin/mailman/listinfo/lldb-commits

Reply via email to