ymandel added a comment. Thanks for expanding the description, including the helpful example. I'm not sure, though, that this is the "right" behavior, at least not always. Worse, I'm not sure there is a single "right" behavior. I can easily imagine a tidy that matches multiple times in the same TU and tries, therefore, to apply the same fix multiple times, even though it wants at most one (and possibly an error). The major problem is that character-level (versus AST level) edits simply don't compose. So, multiple edits on a file are always suspect. Could you also give an example (in terms of code changes, not YAML) of why this comes up? As in, are we sure that the problem lies in the algorithm here, rather than in the phrasing of the transformation itself?
That said, based on some of your linked bugs, it sounds like your change would make the behavior here consistent with the behavior in another situation (when clang-tidy applies the edits directly rather than outputting to YAML?). Consistency could be a strong argument in favor of this change. Repository: rG LLVM Github Monorepo CHANGES SINCE LAST ACTION https://reviews.llvm.org/D76054/new/ https://reviews.llvm.org/D76054 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
