N-Dekker wrote: > You understand that this check does not apply only to std:: containers but > also to boost and other custom one that have size and empty methods. In such > case claim is still valid.
Thanks for your prompt reply, Piotr. I thought of custom containers too, but then again, I would rather not make any claim about the efficiency of `size()` and `empty()` methods of _arbitrary_ custom containers. I mean, they could do anything! In order for the documentation to make claims about `size()` versus `empty()`, I think they should behave like those of std:: containers, including their time complexity. Otherwise it's hard to reason about them in a general way. But then, do you happen to know whether there are indeed still Boost containers that have a `size()` method which has a greater-than-constant time complexity? If so, can you possibly name one? > If you change check description, then check documentation also should be > updated to be in sync. Thanks! So this pull request should make some more adjustments before it can be merged (if it would be accepted at all)? https://github.com/llvm/llvm-project/pull/117629 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org https://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
