[issue9374] urlparse should parse query and fragment for arbitrary schemes

Senthil Kumaran added the comment: Reviewed the issue and correct rollbacks and commits were applied. This ticket should be closed. Thanks! -- status: open -> closed ___ Python tracker __

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Changes by Martin Panter : -- nosy: +vadmium ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: https://mail.python

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Antoine Pitrou added the comment: > It may be fixing a bug, but the bug is not obvious and the fix is not > backward compatible. I therefore suggest to roll back the commits to > 3.2 and 2.7. Well, the bug is quite obvious to me :-) (just hit it here) The fix for those who want the old behaviou

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Roundup Robot added the comment: New changeset 950320c70fb4 by Georg Brandl in branch 'default': Add a versionchanged note for #9374 changes. http://hg.python.org/cpython/rev/950320c70fb4 -- ___ Python tracker _

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Changes by Georg Brandl : -- versions: -Python 3.3 ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http://mail.

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Éric Araujo added the comment: One would hope that people making up URI schemes would follow the generic syntax (and thus irc would be an exception), but as the risk exists I agree we should not break code in bugfix releases. -- ___ Python tracker

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Georg Brandl added the comment: People make up URL schemes all the time, irc:// is not a special case. This change will mean breakage for them, unwarranted. -- ___ Python tracker ___

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Ezio Melotti added the comment: If there is a list of known protocols that don't use the fragment, can't we include it in urlparse as we already do in Lib/urlparse.py:34? If #channel in irc://example.com/#channel should not be parsed as fragment, then this can be considered as a regression. Th

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Georg Brandl added the comment: After encountering an instance of people relying on fragment not being parsed for "irc://" URLs, with resulting breakage, I don't think we should change this in point releases. IOW, it's fine for 3.3.0, but not for 2.7.x or 3.2.x. It may be fixing a bug, but th

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Senthil Kumaran added the comment: Oops. I had not seen Eric and Mattiahs comment to this issue, which pointed out to the problem. Sorry for not acting on this. Thanks Georg for adding those module attributes back. On Fri, Aug 24, 2012 at 9:17 AM, Roundup Robot wrote: > > Roundup Robot added t

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Roundup Robot added the comment: New changeset a0b3cb52816e by Georg Brandl in branch '3.2': Closes #9374: add back now-unused module attributes; removing them is a backward compatibility issue, since they have a public-seeming name. http://hg.python.org/cpython/rev/a0b3cb52816e New changeset c

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Antoine Pitrou added the comment: Senthil, either the module globals should be re-added for compatibility, or the commits should be reverted, IMO. -- nosy: +pitrou ___ Python tracker ___

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Matthias Klose added the comment: this breaks the following upstream builds: createrepo, linkchecker, gwibber, pegasus-wm there is no need to remove is_hierarchical on the branches. it's not used by urlparse at all. is it safe to just keep the uses_query and uses_fragment lists on the branche

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Éric Araujo added the comment: Better link: https://github.com/pypa/pip/issues/552 -- ___ Python tracker ___ ___ Python-bugs-list mail

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Éric Araujo added the comment: Removing the module attributes causes third-party code to break. See one example here: http://lists.idyll.org/pipermail/testing-in-python/2012-July/005082.html -- status: closed -> open ___ Python tracker

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Changes by Arfrever Frehtes Taifersar Arahesis : -- nosy: +Arfrever ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscrib

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Senthil Kumaran added the comment: Thanks for raising this issue, Nick. Yes, I verified in both RFC 3986 and 2396 and realized we can safely adopt a generic parsing system for query and fragment portions of the urls for any scheme. Since it was supported in earlier versions too, I felt it was

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Roundup Robot added the comment: New changeset 79e6ff3d9afd by Senthil Kumaran in branch '2.7': Issue9374 - Generic parsing of query and fragment portion of urls for any scheme http://hg.python.org/cpython/rev/79e6ff3d9afd New changeset a9d43e21f7d8 by Senthil Kumaran in branch '3.2': Issue9374

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Changes by Éric Araujo : -- nosy: +eric.araujo, orsenthil ___ Python tracker ___ ___ Python-bugs-list mailing list Unsubscribe: http:/

[issue9374] urlparse should parse query and fragment for arbitrary schemes

Changes by Ezio Melotti : -- nosy: +ezio.melotti stage: -> needs patch versions: +Python 2.7, Python 3.2, Python 3.3 -Python 2.6 ___ Python tracker ___ __

[issue9374] urlparse should parse query and fragment for arbitrary schemes

New submission from Nick Welch : While the netloc/path parts of URLs are scheme-specific, and urlparse can be forgiven for refusing to parse them for unknown schemes, the query and fragment parts are standardized, and should be parsed for unrecognized schemes. According to Wikipedia: -