On Tue, Aug 9, 2016 at 12:14 PM, Kirill Bobyrev <kbobyrev.opensou...@gmail.com> wrote: > omtcyfz added a comment. > > In https://reviews.llvm.org/D23279#510011, @aaron.ballman wrote: > >> In https://reviews.llvm.org/D23279#510002, @omtcyfz wrote: >> >> > In https://reviews.llvm.org/D23279#509896, @compnerd wrote: >> > >> > > If the argument really is that we want to minimize the tools then Id >> > > argue that `clang-rename` also belongs in `clang-tidy` as it would be >> > > used to rename fields to match the naming convention (tidying up your >> > > code base). >> > >> > >> > It does not belong to `clang-tidy`. `clang-tidy` is a linter, it is meant >> > for diagnosing and fixing typical programming errors. >> >> >> We have the modernize and readability modules which don't really deal with >> programming errors at all (for some definition of programming error). > > > All of the checks in `clang-tidy` **issues**. These are also issues: > > - code style violations > - inefficiency > > And other things. > > There is a clear difference between fixing issues and refactoring. > `clang-format` does refactoring, `clang-tidy` deals with issues.
Consider modernize-loop-convert, modernize-raw-string-literal, and modernize-use-override. These are most definitely refactoring tools that exist within clang-tidy and they do not point out issues with the user's code. That being said, I'm opposed to the proliferation of stand-alone tools when possible. clang-format does not strike me as a good home for the proposed functionality because reformatting does not suggest semantic modifications to the user's source code, and it certainly would not be the first tool I would think to reach for to get this functionality. I think clang-tidy is a more reasonable tool for it. Ideally, we would have a refactoring module that houses most (if not all) of modernize's functionality + refactoring passes that are not really modernizations (such as the recent review of converting memcpy() -> std::copy()). > > > Repository: > rL LLVM > > https://reviews.llvm.org/D23279 > > > _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
