On Mon, Mar 16, 2020 at 11:36 AM Andrew Barnert <abarn...@yahoo.com> wrote:
> > Note that the OP used the word "semantics", but primarily meant > "performance characteristics": > > IIRC, the C++ specification includes guaranteed performance > characteristics as part of its library semantics, Now we are getting into the semantics of the word "semantics" -- beware the RecursionError! > so an implementation whose std::map::find takes more than log N time isn’t > conforming to C++ semantics. I would argue that it's not conforming to the specification, even if it does conform to the semantics :-) > And I don’t think that’s too unreasonable. Nor do I -- a specification should include something about performance characteristics, and we certainly wouldn't want to put anything in the standard library that didn't have reasonable performance for the intended use cases. After all, you CAN just use the existing ones and call sorted() all over the place :-) Anyway, my only point was that we DO want to consider both performance and semantics here -- a Sorted Dict (or SortedList, or ...) has different use cases, and thus should have a different semantics (API). I note that the SortedList, for instance, is not quite a MutableSequence, as it doesn't make sense to support .append() and .extend() for that. Not sure if there is anything about a SortedDict that makes it not a MutableMapping, though. -CHB -- Christopher Barker, PhD Python Language Consulting - Teaching - Scientific Software Development - Desktop GUI and Web Development - wxPython, numpy, scipy, Cython
_______________________________________________ Python-ideas mailing list -- python-ideas@python.org To unsubscribe send an email to python-ideas-le...@python.org https://mail.python.org/mailman3/lists/python-ideas.python.org/ Message archived at https://mail.python.org/archives/list/python-ideas@python.org/message/BPCRV2TV3PGGYUQZYRGW24KZ3SAS6L32/ Code of Conduct: http://python.org/psf/codeofconduct/
