https://github.com/timon-ul requested changes to this pull request.
So I have been looking at `NextToken` and its uses again. What a surprise to find out it is actually related to the `FIXME` you found in `CodeCompleteTests.cpp`. The whole idea of the [commit](https://github.com/llvm/llvm-project/commit/348364bffd379e291501dc49b192cdd2adf83811#diff-4864eebf0abf8b29afb79cfde58b196f40c6baa36e8ad05676325edb08cd577d) is to surpress the creation of additional parenthesis, which in some cases it does, but in others it does not, the issue is tracked [here](https://github.com/clangd/clangd/issues/387) (so you have the already relevant context if you decide to work on that as a follow up). As you can see though, back in that commit the way to get the `NextToken` was by calling `Lexer::findNextToken` which caused other problems why @HighCommander4 actually fixed this behaviour 2 years ago [here](https://github.com/clangd/clangd/issues/1785). All this makes me even more convinced, that the use case of `NextToken` and your reason to create `findEndOfIdentifierContinuation` are heavily overlapping and thus I am convinced we should not duplicate code and calls here, but instad decide on one solution. I am fine by simply adapting your new function for this, but maybe @HighCommander4 could voice his thoughts too given he wrote `findTokenAfterCompletionPoint`. https://github.com/llvm/llvm-project/pull/187623 _______________________________________________ cfe-commits mailing list [email protected] https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
