On Fri, Dec 11, 2009 at 11:23 AM, Nick Coghlan <ncogh...@gmail.com> wrote: [..] > > Eric's suggestion of NormalizedVersion sounds best to me - it exactly > describes the intended role of the class. >
Done. Steve Steiner added a nice functional test that tries the scheme on *all* pypi versions: MacZiade:distutilsver tarek$ nosetests -s . Loading saved pypi data... Results: -------- Total Packages : 8654 Already Match : 7604.0 (87.87%) Have Suggestion : 624.0 (7.21%) No Suggestion : 426.0 (4.92%) ...... ---------------------------------------------------------------------- Ran 6 tests in 0.463s OK IOW, the current scheme works as-is for 88% of the packages, and is able to transform 7% of the remainings. The 5% that are not workable are mostly packages with versions like : (extracts) - .000001 - working proof of concept - release candidate 3 - 1.0dev-BZR-r42-panta-elasticworld.org-20091021153851-6ijlut5dkxndxw1h - beta pre csound - etc.. Furthermore, I've seen some patterns in those 5% that can be worked out so I'll probably be able to lower it to 3% Tarek _______________________________________________ Python-Dev mailing list Python-Dev@python.org http://mail.python.org/mailman/listinfo/python-dev Unsubscribe: http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com