2007-07-23T00:03:54 <dreimark> gn
2007-07-23T00:14:18 <zenhase> re
2007-07-23T00:14:44 <xorAxAx> zenhase: we havent heard anything from fabi yet
2007-07-23T00:16:57 <zenhase> hmm
2007-07-23T00:19:06 <ThomasWaldmann> xorAxAx: i have phoned with him
2007-07-23T00:19:11 <xorAxAx> ah
2007-07-23T00:26:05 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2517:50b12f981890 1.7/MoinMoin/macro/__init__.py: builtin macros: fix bug, handle errors giving help to users
2007-07-23T00:26:06 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2518:fc5b843af8f1 1.7/MoinMoin/parser/ (_tests/test_text_moin_wiki.py text_moin_wiki.py): merge main
2007-07-23T00:26:48 <zenhase> yeah, i did too
2007-07-23T01:01:18 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2519:5f299f9ba77c 1.7/MoinMoin/macro/__init__.py: externalize InterWiki macro, refactor it to use the formatter
2007-07-23T01:07:23 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2520:02766a483465 1.7/MoinMoin/macro/__init__.py: builtin macros: just reorder, no other changes
2007-07-23T02:14:32 <zenhase> need to sleep now
2007-07-23T02:14:56 <zenhase> thomas, xorAxAx: i will work through the list ffesti mailed me
2007-07-23T02:15:45 <ThomasWaldmann> gn zenhase :)
2007-07-23T02:15:53 <grzywacz> My test jabber server died. :<
2007-07-23T02:16:00 <ThomasWaldmann> RIP
2007-07-23T02:16:03 <grzywacz> I guess I must configure a local one, finally. ;)
2007-07-23T02:16:08 <zenhase> grzywacz: try mouth to mouth
2007-07-23T02:16:12 <grzywacz> zenhase, NO :P
2007-07-23T02:16:24 <grzywacz> zenhase, hell knows who was jabbing with it. ;)
2007-07-23T02:16:39 <zenhase> grzywacz: no ssl?
2007-07-23T02:16:46 <zenhase> safer jab
2007-07-23T02:17:07 <grzywacz> zenhase, safer as in "nobody sees you jabbing", but otherwise... :p
2007-07-23T02:18:48 <zenhase> i c
2007-07-23T02:18:58 <zenhase> well, gn8 -> bed
2007-07-23T02:21:58 <grzywacz> night :)
2007-07-23T02:23:42 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2521:b4118a77105d 1.7/MoinMoin/macro/ (__init__.py PageCount.py): externalize PageCount macro
2007-07-23T02:23:44 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2522:d73957833b45 1.7/MoinMoin/macro/InterWiki.py: add InterWiki.py (see cs 5f299f9ba77c)
2007-07-23T02:27:20 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2523:ee8d1a2cc252 1.7/MoinMoin/macro/__init__.py: XSS fix for format_error
2007-07-23T02:33:11 <ThomasWaldmann> heh, TemplateList is undocumented and unused ...
2007-07-23T02:37:59 <grzywacz> lol, it worked...
2007-07-23T02:41:15 <ThomasWaldmann> funny, TemplateList looks similar to 1.3 PageList
2007-07-23T02:43:15 <grzywacz> Now, I have to produce a diff I won't be killed for by xorAxAx. 8)
2007-07-23T02:47:31 <grzywacz> xorAxAx, basically I've just added a new type of notification command, which holds text (to be translated) and a dict of data to interpolate it with after translation.
2007-07-23T02:47:57 <grzywacz> And it just worked with few other changes. O_o;
2007-07-23T02:53:31 <grzywacz> xorAxAx, http://kolos.math.uni.lodz.pl/~grzywacz/argh.patch
2007-07-23T02:53:53 <grzywacz> xorAxAx, initial version for your destructive review. An attempt to i18n xmlrpcbot's messages in xmpp thread
2007-07-23T03:22:13 <grzywacz> GAH
2007-07-23T03:22:23 <grzywacz> dns package used by pyxmpp doesn't resolve "localhost"
2007-07-23T03:39:24 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2524:35dcbfc32a57 1.7/MoinMoin/macro/_tests/test_Hits.py: Fix Hits macro tests
2007-07-23T03:39:26 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2525:ce7fc6ba9b3a 1.7/MoinMoin/_tests/test_sourcecode.py: test_sourcecode: only test recently touched files
2007-07-23T07:31:45 <dreimark> moin
2007-07-23T08:10:03 <dreimark> grzywacz: from 2007-07-21T19:26:22 is test_PageEditor.py already fixed and merged to main?
2007-07-23T08:43:28 <ThomasWaldmann> moin
2007-07-23T08:55:06 <dreimark> ThomasWaldmann: I found a note from xorAxAx in logs from yesterday about test_packages
2007-07-23T08:56:52 <dreimark> Was that failure only introduced by grzywacz broken test?
2007-07-23T09:05:57 <ThomasWaldmann> i didnt analyze the failure or who caused it, but it shows no or a duplicate message (and the tests wants exactly 1 msg)
2007-07-23T09:10:24 <dreimark> looking ...
2007-07-23T09:45:27 <xorAxAx> grzywacz: cant you factor out the whole if not self.token stuff?
2007-07-23T09:46:29 <xorAxAx> grzywacz: i am just wondering about:
2007-07-23T09:46:30 <xorAxAx> - error = _("You must set a (long) secret string!")
2007-07-23T09:46:31 <xorAxAx> + error = "You must set a (long) secret string!"
2007-07-23T10:25:41 <dreimark> attachment dialog [del | move | get | view | install]
2007-07-23T10:26:50 <dreimark> "get" is translated in german into "laden" probably "save as" and "speichern unter" ?
2007-07-23T10:32:21 <dreimark> and author in that line would be nice too (2007-07-23 08:33:17, 0.4 KB)
2007-07-23T10:41:54 <dreimark> grzywacz: page.saveText is broken MoinMoinBugs/1.7dev_page.saveText_utf8_bug
2007-07-23T10:49:25 <dreimark> nearly same line is in __init__ of xmlrpc too
2007-07-23T11:16:02 <dreimark> ThomasWaldmann: xorAxAx: how do i get in git style a changeset (and I do miss the search function)
2007-07-23T11:16:18 <xorAxAx> "get a changeset"?
2007-07-23T11:18:50 <vpv> grzywacz: xmlrpc/__init__.py still does not work, I guess I'll fix it again :P
2007-07-23T11:20:30 <xorAxAx> vpv: write a test
2007-07-23T11:21:06 <dreimark> hmm, I do see 12days ago, 2 weeks ago. the other theme has chown changeset XXXX:FFFAAEE
2007-07-23T11:22:45 <xorAxAx> URL?
2007-07-23T11:24:12 <vpv> you're going to hate this, but I'll push so the fix is available ;)
2007-07-23T11:24:45 <xorAxAx> well, without a test, it may break every next time :)
2007-07-23T11:24:49 <dreimark> vpv: and how could we sure this bug does not occure later again
2007-07-23T11:24:58 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2564:b4c44e5be9e8 1.7-maninfo-vpv/MoinMoin/config/ (_tests/test_multiconfig.py multiconfig.py): password_checker: more builtin checks, more tests
2007-07-23T11:25:19 <vpv> xorAxAx: I guess I could write a test, but I never have, so I'll have to look at some of the examples
2007-07-23T11:25:55 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2565:12854ba50c0d 1.7-maninfo-vpv/MoinMoin/ (config/_tests/test_multiconfig.py userprefs/suid.py): password_checker test: add the usual set import magic for py 2.3, small whitespace fix
2007-07-23T11:26:04 <xorAxAx> vpv: its something that you will love after you have discovered its value :)
2007-07-23T11:26:10 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2566:2dc7a3be8ec9 1.7-maninfo-vpv/MoinMoin/_tests/ (pep8.py test_sourcecode.py): test_sourcecode: check for some silly comparisons with None/True/False
2007-07-23T11:26:18 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2567:70b1a82ec75b 1.7-maninfo-vpv/MoinMoin/_tests/ (pep8.py test_sourcecode.py): test_sourcecode: add pep8 test for 1 space on each side of an operator
2007-07-23T11:26:28 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2568:45641cbe7729 1.7-maninfo-vpv/MoinMoin/ (4 files in 4 dirs): enable new PEP8 test, fixes where it failed
2007-07-23T11:26:38 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2569:c9949c55ff5e 1.7-maninfo-vpv/MoinMoin/ (29 files in 14 dirs): add coverage_modules attribute for coverage testing
2007-07-23T11:26:45 <dreimark> xorAxAx: http://hg.moinmo.in/moin/1.7/rev/a4e8439bf84c utf-8 decoding was introduced. for packages
2007-07-23T11:26:48 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2570:fa3aca6aacd4 1.7-maninfo-vpv/MoinMoin/userprefs/suid.py: make pep8 checker happy
2007-07-23T11:26:54 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2571:3d08f1cb4429 1.7-maninfo-vpv/MoinMoin/ (10 files in 9 dirs): pep8 whitespace fixes
2007-07-23T11:27:04 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2572:ac7f57165c9a 1.7-maninfo-vpv/MoinMoin/userprefs/suid.py: merged main
2007-07-23T11:27:14 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2573:839b360413a9 1.7-maninfo-vpv/MoinMoin/filter/_tests/test_filter.py: tests: add some basic filter tests
2007-07-23T11:27:20 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2574:adf5ec4874ed 1.7-maninfo-vpv/MoinMoin/ (_tests/test_wikiutil.py macro/Hits.py wikiutil.py): fix ParameterParser, rewrite ParameterParser tests for py.test (and add more tests), refactor Hits macro
2007-07-23T11:27:34 <xorAxAx> dreimark: yes, so?
2007-07-23T11:27:36 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2575:18c71299e2bb 1.7-maninfo-vpv/MoinMoin/_tests/test_wikiutil.py: convert test_wikiutil to py.test
2007-07-23T11:27:42 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2576:fe71235f1f7c 1.7-maninfo-vpv/MoinMoin/ (_tests/test_wikiutil.py macro/Hits.py wikiutil.py): ParameterParser: changed to return fixed_count, arg_dict, putting positional args into arg_dict[pos], more tests
2007-07-23T11:27:56 <xorAxAx> indeed, it doesnt show the revno, but its irrelevant anyway
2007-07-23T11:27:58 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2577:15502589ece6 1.7-maninfo-vpv/MoinMoin/ (_tests/test_wikiutil.py wikiutil.py): ParameterParser: check types, remove unused code, tests
2007-07-23T11:28:06 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2578:e0b1fc786797 1.7-maninfo-vpv/MoinMoin/wikiutil.py: ParameterParser: cleanup: avoid using builtins as names, cosmetic changes
2007-07-23T11:28:14 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2579:01f2921e5fa4 1.7-maninfo-vpv/MoinMoin/ (5 files in 5 dirs): tests: convert some more tests to py.test
2007-07-23T11:28:24 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2580:8ae20e588337 1.7-maninfo-vpv/MoinMoin/action/_tests/test_attachfile.py: Add a test for AttachFile action.
2007-07-23T11:28:32 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2581:99b6222544c3 1.7-maninfo-vpv/MoinMoin/userprefs/prefs.py: Fix a silly bug in userprefs thanks to johill.
2007-07-23T11:28:38 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2582:f456dc2048d1 1.7-maninfo-vpv/ (8 files in 4 dirs): i18n should now work. First step, no caching yet.
2007-07-23T11:28:46 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2583:8ce2afa469a7 1.7-maninfo-vpv/ (9 files in 7 dirs): Merge main.
2007-07-23T11:28:50 <dreimark> xorAxAx: http://moinmoin.wikiwikiweb.de/MoinMoinBugs/1.7dev_page.saveText_utf8_bug
2007-07-23T11:28:54 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2584:920089399dfa 1.7-maninfo-vpv/MoinMoin/i18n/jabberbot/pl.JabberBot.po: Initial Polish translation for the notification bot.
2007-07-23T11:29:02 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2585:3c5b329afe4b 1.7-maninfo-vpv/MoinMoin/events/__init__.py: Add a pseudo-getText function to that strings can get into .po files
2007-07-23T11:29:08 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2586:474b7a1fc031 1.7-maninfo-vpv/MoinMoin/ (14 files in 7 dirs): Merge main.
2007-07-23T11:29:16 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2587:474448e88c24 1.7-maninfo-vpv/MoinMoin/ (14 files in 7 dirs): Merge main.
2007-07-23T11:29:16 <dreimark> xorAxAx: änderung (don't use ä)
2007-07-23T11:29:24 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2588:31b3d03bd1ea 1.7-maninfo-vpv/MoinMoin/events/__init__.py: s/An/A/
2007-07-23T11:29:30 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2589:4a199a4c5d31 1.7-maninfo-vpv/MoinMoin/userprefs/notification.py: Pull some more johill's changes from main.
2007-07-23T11:29:37 <CIA-27> (59 lines omitted)
2007-07-23T11:31:54 <xorAxAx> dreimark: the file is invalid
2007-07-23T11:32:01 <xorAxAx> dreimark: it must be encoded using utf-8
2007-07-23T11:32:02 <dreimark> why
2007-07-23T11:32:08 <xorAxAx> by definition
2007-07-23T11:32:14 <xorAxAx> --> its not a bug
2007-07-23T11:32:51 <dreimark> ok
2007-07-23T11:33:37 <vpv> so here's the fix, this has been the bug all the time: http://hg.moinmo.in/moin/1.7-maninfo-vpv/rev/aa428c7444e3
2007-07-23T11:33:38 <johill> dreimark: your pre stuff is borked
2007-07-23T11:34:21 <xorAxAx> johill: write a test :)
2007-07-23T11:34:25 <dreimark> johill: ?
2007-07-23T11:34:41 <xorAxAx> johill may backport http://hg.moinmo.in/moin/1.7-maninfo-vpv/rev/aa428c7444e3
2007-07-23T11:34:42 <johill> I did :)
2007-07-23T11:34:42 <johill> about to ococommit it
2007-07-23T11:34:43 <xorAxAx> :-)
2007-07-23T11:35:00 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2526:860e7bf6c4cc 1.7/MoinMoin/parser/_tests/test_text_moin_wiki.py: add failing {{{ }}} test
2007-07-23T11:36:29 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2527:4f6f7e56eb40 1.7/MoinMoin/wikiutil.py: fix error message for get_* (arg parser)
2007-07-23T11:36:59 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2528:1ae621d83c5b 1.7/MoinMoin/ (_tests/test_wikiutil.py wikiutil.py): fix arg parser bug, test for this bug
2007-07-23T11:39:16 <johill> I had another idea for the arg parser
2007-07-23T11:39:49 <johill> look: http://johannes.sipsolutions.net/patches/moin/macro-invoke-defaults.patch
2007-07-23T11:40:15 <johill> only bad thing is that it uses fn.func_code
2007-07-23T11:40:20 <vpv> I think I'm getting the idea of making those tests, I can make a really simple test for that one case and then it can maybe be expanded if needed
2007-07-23T11:42:03 <johill> can we rely on the internals of code objects? :)
2007-07-23T11:42:47 <dreimark> johill: the difference to 1.5 is only <span class="anchor" id="line-0"> before pqr
2007-07-23T11:42:51 <xorAxAx> johill: yes
2007-07-23T11:43:26 <johill> dreimark: no no no
2007-07-23T11:43:38 <johill> dreimark: in 1.6/1.7 you have one pre section for both, not two tt
2007-07-23T11:43:42 <dreimark> did you have seen my patch from yesterday
2007-07-23T11:43:51 <johill> oh, when yesterday?
2007-07-23T11:44:06 <johill> ah
2007-07-23T11:44:08 <johill> nope :)
2007-07-23T11:44:14 <johill> I stand corrected
2007-07-23T11:44:18 <johill> mind fixing the test then?
2007-07-23T11:44:38 <dreimark> changeset 2516 05c9d52dcf5d
2007-07-23T11:44:58 <johill> yeah just sfound it
2007-07-23T11:45:10 <johill> I still think this nested pre thing is pretty borked tho
2007-07-23T11:45:15 <dreimark> it is
2007-07-23T11:45:17 <johill> because it has to nest perfectly
2007-07-23T11:45:25 <dreimark> it is borked
2007-07-23T11:45:33 <johill> have you looked how wiki creole does it?
2007-07-23T11:45:43 <dreimark> nope
2007-07-23T11:45:57 <johill> first of all they require {{{ and }}} on own lines
2007-07-23T11:46:05 <johill> without any space before
2007-07-23T11:46:30 <johill> and if you want }}} inside you quote it with a space in front which is removed
2007-07-23T11:46:43 <johill> a la re.compile("^ }}}")
2007-07-23T11:47:05 <johill> I suppose if you want to allow nesting you could do the same thing with {{{ as well
2007-07-23T11:47:21 <dreimark> yeah but this is a syntax change and we have had for SyntaxReference some discussions
2007-07-23T11:47:37 <dreimark> because of * {{{source}}}
2007-07-23T11:47:55 <johill> yeah
2007-07-23T11:48:05 <johill> we could relax those rules a bit to apply only when you want nesting
2007-07-23T11:48:48 <dreimark> I will not have time do work on that, because I will leave on wednesday
2007-07-23T11:49:19 <dreimark> and I need some vacation
2007-07-23T11:49:45 <johill> ok
2007-07-23T11:51:34 <dreimark> while we do parse line by line I thought counting {{{ and counting }}} would be a good idea
2007-07-23T11:52:55 <dreimark> but later on i recognized that it will fail using it in one line.
2007-07-23T11:53:39 <dreimark> so currently only the SyntaxReference and similiar is ok
2007-07-23T11:54:22 <johill> dunno
2007-07-23T11:54:35 <johill> I just immensely dislike the code doing all this ;)
2007-07-23T11:55:06 <johill> anyway, I wanted to explain that patch too :)
2007-07-23T11:55:36 <johill> the patch I just showed basically converts all macro arguments to the type of the default value of that argument
2007-07-23T11:55:57 <johill> so if you have a macro_xx(macro, a=1) then you won't ever get any non-int/long for a
2007-07-23T11:56:15 <johill> and thus you don't need to call get_int()
2007-07-23T11:56:18 <johill> oh btw
2007-07-23T11:56:48 <johill> xorAxAx: I know, it's not really nice, but do you have a good idea on how to make the number of args fit? unless you want to parse the string and recreate it with count-1 ...
2007-07-23T11:59:55 <dreimark> johill: I would prefer a syntax change of the parsers code for pre too.
2007-07-23T12:01:42 <xorAxAx> johill: :)
2007-07-23T12:02:09 <xorAxAx> johill: its indeed not easy
2007-07-23T12:03:42 <johill> can we rely on the string to be the same on all python versions? or should we maybe generate the string on startup by forcing that exception :)
2007-07-23T12:03:42 <xorAxAx> johill: you could construct a wrapper using eval() that has the same signature like the to be called function but gets the macro object via the closure cell
2007-07-23T12:04:13 <xorAxAx> but you would need some way to get the old signature easily as a string
2007-07-23T12:04:38 <xorAxAx> but maybe hacking the exception is easier
2007-07-23T12:04:45 <johill> getting the signature is possible
2007-07-23T12:04:50 <xorAxAx> sure
2007-07-23T12:05:02 <johill> you can get the arg names, defaults, and flags whether it ias * or **
2007-07-23T12:05:12 <johill> but it's quite a bit of hacking around with that stuff :)
2007-07-23T12:05:20 <xorAxAx> yes, not sure if its worth the effort :)
2007-07-23T12:05:21 <johill> but how does that closure thing work anyway?
2007-07-23T12:05:24 <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2649:6b4fc04d700d 1.7-maninfo-vpv/MoinMoin/xmlrpc/_tests/test_xmlrpc.py: add test for xmlrpc getAuthToken
2007-07-23T12:05:43 <xorAxAx> well, when compiling the new function, python binds the macro name to the macro object in the caller frame
2007-07-23T12:06:17 <vpv> there's a test, it's really simple since it uses a dummy uid/pass pair, but it does catch the bug I hit
2007-07-23T12:06:56 <vpv> http://hg.moinmo.in/moin/1.7-maninfo-vpv/rev/6b4fc04d700d
2007-07-23T12:07:27 <xorAxAx> def call_macro(name, args): *args, **kwargs = do_parse(args)\nmacro=object()\nf_to_call=load_macro(name)\nl={}\neval("....",l)\nf=l["f"]\nf(*args, **kwargs)
2007-07-23T12:07:32 <xorAxAx> something along the lines of that :)
2007-07-23T12:07:34 <johill> ugh :)
2007-07-23T12:08:03 <johill> I think I'd rather do the exception message hacking
2007-07-23T12:08:17 <xorAxAx> hehe
2007-07-23T12:10:53 <grzywacz> vpv, what does not work?
2007-07-23T12:11:08 <grzywacz> dreimark, what's broken in saveText?
2007-07-23T12:11:32 <dreimark> grzywacz: nothing I was wrong
2007-07-23T12:11:36 <vpv> grzywacz: see http://hg.moinmo.in/moin/1.7-maninfo-vpv/rev/aa428c7444e3 and http://hg.moinmo.in/moin/1.7-maninfo-vpv/rev/6b4fc04d700d
2007-07-23T12:11:55 <dreimark> grzywacz: input has to be utf-8
2007-07-23T12:12:40 <vpv> grzywacz: feel free to improve the tests, especially if you know how to call getAuthToken with an existing uid/pass for the test wiki
2007-07-23T12:13:22 <grzywacz> erm
2007-07-23T12:14:46 <grzywacz> Why do you have "self.u = blah" in there in the first place?
2007-07-23T12:15:08 <grzywacz> It's just "u = blah()" in my code <_<
2007-07-23T12:17:36 <dreimark> johill: http://docs.python.org/lib/inspect-classes-functions.html getargspec( func)
2007-07-23T12:20:03 <vpv> grzywacz: yes and it crashes
2007-07-23T12:21:19 <vpv> grzywacz: because you store the handle_auth result to self.u and then call u later, u is not defined, exception
2007-07-23T12:21:54 <grzywacz> I don't store anything in self.u, there's no self.u in my code :E
2007-07-23T12:23:43 <vpv> you're correct, there isn't, this must be a merge error or something
2007-07-23T12:24:24 <xorAxAx> hg annotate
2007-07-23T12:25:08 <ThomasWaldmann> johill: nice hack :) and it is even possible to get unconverted stuff with default=None afaics.
2007-07-23T12:26:18 <johill> ThomasWaldmann: sure
2007-07-23T12:26:26 <johill> want it?
2007-07-23T12:26:40 <johill> ah getargspec looks good
2007-07-23T12:26:42 <johill> thanks I'll use that
2007-07-23T12:26:46 <ThomasWaldmann> btw, one reason why I was using get_<type> is that it enables you to raise useful ValueErrors that get displayed in a nice way to the wiki user
2007-07-23T12:27:06 <johill> right but I use them too internally
2007-07-23T12:27:13 <johill> (and just fixed up the error messages to be more verbose)
2007-07-23T12:29:12 <ThomasWaldmann> yeah, commit it (if we can be rather sure this works for py 2.x, x>=3)
2007-07-23T12:29:55 <johill> yeah I need to see 2.3
2007-07-23T12:30:13 <johill> do they have the old docs online?
2007-07-23T12:30:48 <johill> yeah docs say it has getargspec
2007-07-23T12:31:18 <dreimark> :)
2007-07-23T12:31:32 <johill> thanks :)
2007-07-23T12:32:05 <ThomasWaldmann> please also add some docstrings/comments, this is stuff most py programmers are not used to
2007-07-23T12:32:07 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2529:df375ab9fbde 1.7/MoinMoin/macro/__init__.py: automatically convert macro args with defaults to the default's type
2007-07-23T12:32:22 <johill> I think I should just put it into the wiki docs
2007-07-23T12:32:33 <johill> or put it into the macro module docstring?
2007-07-23T12:32:50 <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2650:b41c924c4f4c 1.7-maninfo-vpv/MoinMoin/xmlrpc/__init__.py: fix some old weirdness causing bugs in xmlrpc init
2007-07-23T12:33:44 <ThomasWaldmann> I meant programmer's. For the users, we need a macro writing howto for 1.7.
2007-07-23T12:33:53 <vpv> xorAxAx: I think I won't start playing with annotate, it's probably me who has messed up some merge, it should be alright now
2007-07-23T12:34:16 <vpv> But if you want to take my test to main, feel free
2007-07-23T12:34:20 <johill> ThomasWaldmann: oh ok
2007-07-23T12:34:50 <xorAxAx> vpv: messing up a merge is fatal
2007-07-23T12:34:59 <xorAxAx> vpv: IMHO you should analyse what went wrong
2007-07-23T12:35:23 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2530:0a8fc701e40d 1.7/MoinMoin/macro/__init__.py: use the new automatic macro arg conversion in some internal macros
2007-07-23T12:43:06 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2531:1184922dc419 1.7/MoinMoin/macro/__init__.py: some more explanations of the type magic for macro args
2007-07-23T12:43:32 <johill> I guess that all the code I just put there should be ported into wikiutil when first needed elsewhere (parser args?)
2007-07-23T12:45:00 <ThomasWaldmann> johill: shouldn't have every macro arg have a default now (sometimes empty wanted type or None in "don't touch" cases)?
2007-07-23T12:45:31 <johill> not necessarily
2007-07-23T12:45:42 <johill> if you don't have one then that argument is required
2007-07-23T12:46:18 <vpv> xorAxAx: I think I found the reason, http://hg.moinmo.in/moin/1.7-maninfo-vpv/rev/5fdc953c8961
2007-07-23T12:46:23 <ThomasWaldmann> but you wont get type conversion then
2007-07-23T12:47:29 <johill> right
2007-07-23T12:47:58 <xorAxAx> vpv: ok
2007-07-23T12:48:09 <xorAxAx> johill: did you port 5fdc953c8961 to main?
2007-07-23T12:48:21 <xorAxAx> johill: then you need to do so with b41c924c4f4c as well
2007-07-23T12:48:55 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2532:cc125871c8a0 1.7/MoinMoin/macro/__init__.py: whoops, get_boolean was renamed to get_bool
2007-07-23T12:49:04 <johill> xorAxAx: no, not yet
2007-07-23T12:49:29 <ThomasWaldmann> johill: for regularity, we could just tell "give defaults anywhere" and that macro authors have to check emtyness (or none) in the macro code
2007-07-23T12:49:36 <vpv> grzywacz: so I fixed an earlier bug by using self.u myself (as can be seem from that link) which caused a different bug later...
2007-07-23T12:51:02 <ThomasWaldmann> the only special case then would be giving None as default as "no convert" indicator
2007-07-23T12:51:18 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2533:19ba0aa5ccdf 1.7/MoinMoin/parser/_tests/test_text_moin_wiki.py: test_text_moin_wiki:fixed output from testMultipleShortPreSections
2007-07-23T12:51:19 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2534:4fb5d4b43d4c 1.7/MoinMoin/macro/__init__.py: branch merge
2007-07-23T12:51:22 <johill> ThomasWaldmann: that'd also get rid of the 'invalid number of arguments' message
2007-07-23T12:51:36 <vpv> johill, xorAxAx: the only changeset worth pushing to main might be http://hg.moinmo.in/moin/1.7-maninfo-vpv/rev/6b4fc04d700d at this point...
2007-07-23T12:51:46 <ThomasWaldmann> right :) so we don't have to do surgery there
2007-07-23T12:51:53 <xorAxAx> vpv: one was committed
2007-07-23T12:52:04 <johill> ThomasWaldmann: however, if you actually need the first two args you need to check the error yourself...
2007-07-23T12:53:05 <dreimark> johill: one test fails (test_text_moin_wiki.py) TypeError: args must be unicode
2007-07-23T12:53:08 <ThomasWaldmann> yes, but you can just do if not x or not y: raise ValueError("x,y params missing")
2007-07-23T12:53:31 <ThomasWaldmann> hmm, difficult for int
2007-07-23T12:54:01 <vpv> xorAxAx: nothing of mine has been pushed yet, or I can't see it
2007-07-23T12:54:06 <johill> dreimark: hm?
2007-07-23T12:54:09 <xorAxAx> vpv: ask johill
2007-07-23T12:54:54 <johill> dreimark: ah I see
2007-07-23T12:54:59 <ThomasWaldmann> johill: it would be nice to have macro_square(x=int)
2007-07-23T12:55:11 <johill> dreimark: well, bogus test, wiki markup input should also be a unicode object
2007-07-23T12:55:15 <johill> ThomasWaldmann: as an example?
2007-07-23T12:55:33 <ThomasWaldmann> so the conversion will now int is wanted, but it will still use None if there was nothing
2007-07-23T12:55:57 <johill> ThomasWaldmann: oh you mean give the type as default instead of a value?
2007-07-23T12:56:10 <ThomasWaldmann> alternatively, yes
2007-07-23T12:56:23 <johill> macro_square(x=int) looked like that
2007-07-23T12:56:32 <ThomasWaldmann> yes, the type
2007-07-23T12:57:01 <ThomasWaldmann> otherwise you could not give any useful default
2007-07-23T12:57:06 <johill> right
2007-07-23T12:57:07 <johill> makes sense
2007-07-23T12:57:45 <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2651:849142a76ea1 1.7-maninfo-vpv/MoinMoin/xmlrpc/__init__.py: fix _cleanup_stale_tokens call
2007-07-23T12:57:47 <johill> dreimark: fixed
2007-07-23T12:57:47 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2535:52961e122768 1.7/MoinMoin/parser/_tests/test_text_moin_wiki.py: fix tests, using unicode instead of str
2007-07-23T12:57:55 <ThomasWaldmann> so we have donttouch=None, x=int, text=u''
2007-07-23T12:57:59 <johill> right
2007-07-23T12:58:58 <ThomasWaldmann> (of course one could also use text=unicode if you want to have real empty values and differentiate them from the default, which is None then)
2007-07-23T12:59:11 <vpv> johill: as far as I can see, these two might be worth taking to main: http://hg.moinmo.in/moin/1.7-maninfo-vpv/rev/6b4fc04d700d and http://hg.moinmo.in/moin/1.7-maninfo-vpv/rev/849142a76ea1
2007-07-23T12:59:53 <vpv> the latter should be a real bugfix, the bug only happens once in a 100 calls, so it might be though to catch
2007-07-23T12:59:57 <ThomasWaldmann> johill: you could suggest that stuff as a PEP :D
2007-07-23T13:00:10 <johill> hah not really
2007-07-23T13:00:17 <ThomasWaldmann> (put on asbestos underwear first)
2007-07-23T13:00:20 <johill> you might actually want a function that takes types as input
2007-07-23T13:07:38 <johill> works
2007-07-23T13:09:09 <ThomasWaldmann> johill: remove that TypeError catching then?
2007-07-23T13:09:30 <johill> not sure
2007-07-23T13:09:38 <johill> it's still nicer if somebody has a bug instead of getting a backtrace
2007-07-23T13:09:57 <johill> and I know most people only test the good cases, not the error cases
2007-07-23T13:11:03 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2536:72a31f09d5a1 1.7/MoinMoin/macro/__init__.py: allow types as defaults for macro args to force conversion to that type
2007-07-23T13:12:15 <johill> of course, if you give a type and the user skips over that arg by ,,
2007-07-23T13:12:16 <dreimark> macro_Hits(filter=('VIEWPAGE', 'SAVEPAGE')) (OR) some advanced logic possible too?
2007-07-23T13:12:18 <johill> then you still get None
2007-07-23T13:12:44 <johill> dreimark: uh huh, yes, it could be possible but I think we're approaching the "went too far" line
2007-07-23T13:12:57 <johill> :)
2007-07-23T13:14:44 <ThomasWaldmann> johill: yeah, None is wanted in that case
2007-07-23T13:15:14 <johill> yup, that's what you get
2007-07-23T13:15:26 <johill> but if it's int then you get None or an int
2007-07-23T13:15:32 <johill> instead of None or a unicode object
2007-07-23T13:15:35 <johill> etc
2007-07-23T13:15:56 <ThomasWaldmann> fine
2007-07-23T13:16:16 <dreimark> would be nice for refactoring of http://moinmoin.wikiwikiweb.de/ParserMarket?action=AttachFile&do=view&target=text_x_frame-1.6.0-5.py
2007-07-23T13:17:04 <dreimark> all similiar lines if self.align in ['left', 'right', 'center', 'justify']: could be removed
2007-07-23T13:17:21 <johill> I guess
2007-07-23T13:17:45 <johill> I suppose it'd be good for translation
2007-07-23T13:17:54 <johill> let me just write the wiki text explaining it all
2007-07-23T13:18:18 <ThomasWaldmann> if you extend that help page, list new method first
2007-07-23T13:20:42 <johill> I'm actually removing the old method
2007-07-23T13:27:49 <johill> I wonder if we should require arguments that have no default
2007-07-23T13:27:58 <johill> instead of allowing to pass None by skipping over
2007-07-23T13:28:45 <johill> not sure tho, probably not
2007-07-23T13:29:15 <johill> ThomasWaldmann: http://moinmoin.wikiwikiweb.de/MoinMoinTodo/Release_1.7/HelpOnMacros#head-978aab0a4ec3948d1a9b81efd19e8ac014653450
2007-07-23T13:30:50 <johill> dreimark: not too bad, but what about case-sensitivity vs. case-insensitivity?
2007-07-23T13:35:11 <johill> ah
2007-07-23T13:35:22 <johill> dreimark: no, it's not really easy to do because it would mean appending args that weren't given
2007-07-23T13:36:53 <johill> and that collides with keyword args
2007-07-23T13:37:10 <dreimark> johill: it can get too complicated.
2007-07-23T13:38:16 <johill> would could add a helper
2007-07-23T13:38:19 <johill> get_choice()
2007-07-23T13:38:39 <ThomasWaldmann> johill: 1. [[Example()]] - error, not enough arguments isnt that irregular?
2007-07-23T13:38:49 <johill> why?
2007-07-23T13:39:05 <ThomasWaldmann> shouldnt it give None, None?
2007-07-23T13:39:20 <johill> hm
2007-07-23T13:39:35 <ThomasWaldmann> (compare bottom example with the one right above)
2007-07-23T13:39:54 <johill> right
2007-07-23T13:40:08 <johill> that gets into the hairy case dreimark suggested
2007-07-23T13:40:25 <johill> imagine I add [[Example(arg2=7)]]
2007-07-23T13:40:40 <johill> maybe I shouldn't use positional args at all
2007-07-23T13:40:44 <johill> but convert everything to keyword
2007-07-23T13:40:48 <johill> I have the varnames anyway
2007-07-23T13:40:51 <johill> and then this all gets easier
2007-07-23T13:40:59 <johill> I'll do that later, need to have lunch now
2007-07-23T13:41:10 <dreimark> bon appetite
2007-07-23T13:41:25 <ThomasWaldmann> hasta la pasta :)
2007-07-23T13:41:49 <johill> I hope that works anyway
2007-07-23T13:42:10 <johill> yeah should
2007-07-23T13:42:14 <johill> well, you'll see tomorrow :)
2007-07-23T13:43:08 <ThomasWaldmann> :)
2007-07-23T13:43:16 <dreimark> :)
2007-07-23T13:44:51 <ThomasWaldmann> bbl
2007-07-23T13:58:36 <vpv> I hope you could be the fastest to answer or give hints than a random Google search to this question: How could I import different things for different Python versions?
2007-07-23T13:59:26 <xorAxAx> vpv: what do you want to import?
2007-07-23T13:59:33 <vpv> For example, now that I have Python 2.5 myself, cElementTree doesn't work as such, since it's probably under xml. In 2.4 it was an extra module.
2007-07-23T13:59:50 <vpv> (outside core Python, that is)
2007-07-23T14:00:01 <xorAxAx> vpv: for elementtree, you need the famous 10 lines import statement
2007-07-23T14:00:17 <xorAxAx> http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/475126
2007-07-23T14:00:26 <xorAxAx> oh, 15 lines
2007-07-23T14:01:37 <vpv> omg :D but try - except is generally the good practice here
2007-07-23T14:01:57 <vpv> I'll come up with something, I'd rather use cElementTree if possible
2007-07-23T14:02:03 <vpv> thanks
2007-07-23T14:07:01 <ThomasWaldmann> vpv: you need this for what?
2007-07-23T14:07:18 <xorAxAx> good question
2007-07-23T14:07:41 <vpv> xml parsing :D
2007-07-23T14:07:47 <ThomasWaldmann> hah
2007-07-23T14:07:53 <xorAxAx> well, what do you want to parse and why?
2007-07-23T14:08:02 <vpv> this is again Fedora specific, the repository metadata is in XML
2007-07-23T14:08:21 <vpv> and the files can be something like 6 megs or more in size
2007-07-23T14:11:52 <vpv> But I'll do the import hack only in the function where I need it, so it won't affect other possible users/usecases
2007-07-23T14:19:52 <xorAxAx> ah
2007-07-23T14:25:24 <dreimark> ThomasWaldmann: simple password check missed to test for reversed written username
2007-07-23T14:28:17 <ThomasWaldmann> dreimark: feel free to add code and test :)
2007-07-23T14:28:27 <dreimark> hehe
2007-07-23T14:31:08 <dreimark> will do
2007-07-23T14:33:59 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2537:b257d7bc96ff 1.7/MoinMoin/action/AttachFile.py: AttachFile:only superuser with write access to a page should get install dialog for packages
2007-07-23T14:36:01 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2652:860e7bf6c4cc 1.7-maninfo-vpv/MoinMoin/parser/_tests/test_text_moin_wiki.py: add failing {{{ }}} test
2007-07-23T14:36:03 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2653:4f6f7e56eb40 1.7-maninfo-vpv/MoinMoin/wikiutil.py: fix error message for get_* (arg parser)
2007-07-23T14:36:06 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2654:1ae621d83c5b 1.7-maninfo-vpv/MoinMoin/ (_tests/test_wikiutil.py wikiutil.py): fix arg parser bug, test for this bug
2007-07-23T14:36:12 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2655:df375ab9fbde 1.7-maninfo-vpv/MoinMoin/macro/__init__.py: automatically convert macro args with defaults to the default's type
2007-07-23T14:36:25 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2656:0a8fc701e40d 1.7-maninfo-vpv/MoinMoin/macro/__init__.py: use the new automatic macro arg conversion in some internal macros
2007-07-23T14:36:31 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2657:1184922dc419 1.7-maninfo-vpv/MoinMoin/macro/__init__.py: some more explanations of the type magic for macro args
2007-07-23T14:36:39 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2658:cc125871c8a0 1.7-maninfo-vpv/MoinMoin/macro/__init__.py: whoops, get_boolean was renamed to get_bool
2007-07-23T14:36:47 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2659:19ba0aa5ccdf 1.7-maninfo-vpv/MoinMoin/parser/_tests/test_text_moin_wiki.py: test_text_moin_wiki:fixed output from testMultipleShortPreSections
2007-07-23T14:36:57 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2660:4fb5d4b43d4c 1.7-maninfo-vpv/MoinMoin/macro/__init__.py: branch merge
2007-07-23T14:37:03 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2661:52961e122768 1.7-maninfo-vpv/MoinMoin/parser/_tests/test_text_moin_wiki.py: fix tests, using unicode instead of str
2007-07-23T14:37:13 <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2662:ada5d90e0142 1.7-maninfo-vpv/MoinMoin/ (4 files in 4 dirs): merge from main
2007-07-23T14:37:21 <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2663:196a7494d517 1.7-maninfo-vpv/MoinMoin/script/xmlrpc/manimport.py: Add cElementTree 2.4 vs 2.5 tricks, fix one small bug
2007-07-23T14:37:29 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2664:72a31f09d5a1 1.7-maninfo-vpv/MoinMoin/macro/__init__.py: allow types as defaults for macro args to force conversion to that type
2007-07-23T14:37:35 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2665:b257d7bc96ff 1.7-maninfo-vpv/MoinMoin/action/AttachFile.py: AttachFile:only superuser with write access to a page should get install dialog for packages
2007-07-23T14:37:43 <CIA-27> moin: Ville-Pekka Vainio <vpivaini AT cs DOT helsinki DOT fi> * 2666:da96ec31ad6a 1.7-maninfo-vpv/MoinMoin/ (action/AttachFile.py macro/__init__.py): merge from main
2007-07-23T14:38:15 <dreimark> bbl, power off
2007-07-23T15:14:42 <lanius> ThomasWaldmann: did you take a deeper look at my code?
2007-07-23T15:24:37 <ThomasWaldmann> not yet (except reading diffs, of course), but I did some practical tests (and posted the results here some days ago)
2007-07-23T15:25:03 <ThomasWaldmann> how's your indexing stuff going?
2007-07-23T15:25:07 <ThomasWaldmann> lanius:
2007-07-23T15:25:15 <lanius> ThomasWaldmann: the indexing stuff is done
2007-07-23T15:26:00 <ThomasWaldmann> so the edit-log is still missing, right?
2007-07-23T15:26:03 <xorAxAx> lanius: please refactor all usages of meta-data keys to use common CONSTANTS ":)"
2007-07-23T15:26:09 <lanius> ThomasWaldmann: right
2007-07-23T15:26:16 <lanius> xorAxAx: where is it missing
2007-07-23T15:26:29 <xorAxAx> lanius: there where you added DELETED lately
2007-07-23T15:26:40 <xorAxAx> the list has only strings in it besides DELETEDD
2007-07-23T15:27:21 <lanius> xorAxAx: well, hmm, those are not my metadta keys ;)
2007-07-23T15:27:32 <xorAxAx> aha?
2007-07-23T15:28:11 <lanius> well, i just added them to be an index, they where there before
2007-07-23T15:28:32 <xorAxAx> yeah, but what are they if not keys?
2007-07-23T15:28:44 <lanius> they are keys, just not "my keys" ;)
2007-07-23T15:29:08 <xorAxAx> your keys?
2007-07-23T15:29:25 <lanius> keys that have been there before
2007-07-23T15:29:46 <xorAxAx> moin didnt have metadata before
2007-07-23T15:29:49 <xorAxAx> so it cant have any keys
2007-07-23T15:29:58 <lanius> user attributes
2007-07-23T15:30:02 <xorAxAx> ?
2007-07-23T15:30:16 <lanius> email, jid, names, openids are attributes of an user
2007-07-23T15:30:21 <xorAxAx> ah, well
2007-07-23T15:30:24 <lanius> which before was no idea what and now are metadata kesy
2007-07-23T15:30:39 <xorAxAx> are you handling the mime type search problem?
2007-07-23T15:30:55 <lanius> no, moved that to extra functionality
2007-07-23T15:31:02 <lanius> i.e. later
2007-07-23T15:31:10 <xorAxAx> ok
2007-07-23T15:31:17 <lanius> but i don't see issues integrating it
2007-07-23T15:31:22 <xorAxAx> me neither
2007-07-23T15:36:39 <lanius> xorAxAx: ThomasWaldmann i still need a good idea how to properly implement the edit-log
2007-07-23T15:37:23 <lanius> if i implement it as metadata for the items we loose the global edit-log functionality
2007-07-23T15:37:34 <lanius> and the performance might be really bad for the current storage
2007-07-23T15:37:37 <grzywacz> So...
2007-07-23T15:38:00 <grzywacz> Is there a chance that page comments will be separate from page text in lanius' branch?
2007-07-23T15:38:43 <lanius> page comments are stored in the edit-log currently, right?
2007-07-23T15:39:19 <grzywacz> No, not comments related to changes.
2007-07-23T15:39:29 <lanius> what kind of comments then
2007-07-23T15:39:30 <grzywacz> More like :Talk pages in mediawiki.
2007-07-23T15:39:40 <lanius> don't know that
2007-07-23T15:39:54 <xorAxAx> grzywacz: thats a matter of theme
2007-07-23T15:39:58 <grzywacz> Basically, I'm asking it will be possible to add arbitrary data to a page, without altering its text?
2007-07-23T15:40:05 <grzywacz> xorAxAx, hrmpf. Is it?
2007-07-23T15:40:51 <xorAxAx> grzywacz: yes
2007-07-23T15:41:00 <grzywacz> Ok...
2007-07-23T15:41:12 <xorAxAx> the new data model doesnt support orthogonal data that is meant to be displayed to the user
2007-07-23T15:41:19 <xorAxAx> of course you could store comments as metadata, though
2007-07-23T15:43:09 <lanius> xorAxAx: orthogonal is one of your favorite words, isn't it :-)
2007-07-23T15:43:27 <xorAxAx> lanius: yes :)
2007-07-23T15:43:49 <xorAxAx> lanius: in a few months, it wont be anymore i am sure
2007-07-23T15:44:02 <lanius> xorAxAx: you could write a book about it
2007-07-23T15:44:55 <lanius> well, any idea about the edit-log stuff
2007-07-23T15:44:58 <xorAxAx> orthogonal orthogonalities of orthogonal system properties in orthogonal cybernetic systems
2007-07-23T15:45:01 <xorAxAx> no
2007-07-23T15:45:23 * xorAxAx is currently in exam learning mode
2007-07-23T15:45:27 <xorAxAx> dont you have any exams?
2007-07-23T15:45:31 <lanius> looking at how it is realized in other sucking wiki's might be a good idea
2007-07-23T15:45:41 <lanius> xorAxAx: no my exams are over, i'm in working mode
2007-07-23T15:45:56 <xorAxAx> hehe
2007-07-23T15:46:02 <lanius> that means real work besides moin :-)
2007-07-23T15:46:16 <xorAxAx> so you should be in the company today?
2007-07-23T15:46:22 <lanius> that's where i am
2007-07-23T15:46:29 <xorAxAx> lanius: well, i am sure that other wikis simply have a table that contains a list of changes
2007-07-23T15:46:34 <xorAxAx> with a timestamp
2007-07-23T15:46:52 <lanius> this would end up in another interface at the storage layer, which i already consider
2007-07-23T15:47:20 <xorAxAx> IMHO the new interface shouldnt be orthogonal but tightly integrated into the storage backend :)
2007-07-23T15:47:55 <xorAxAx> make a proposal in the wiki maybe
2007-07-23T15:47:57 <lanius> making it metadata would be tightly integrated, but it would be missing the global edit-log functionality then
2007-07-23T15:48:34 <lanius> ThomasWaldmann: any idea from your side?
2007-07-23T15:48:48 <xorAxAx> how about having an item where the metadata is used for the edit-log, revision wise?
2007-07-23T15:49:07 <xorAxAx> so you just need to read the last n metadata revisions to display recentchanges
2007-07-23T15:49:08 <ThomasWaldmann> just a moment, i am discussing with zenhase right now
2007-07-23T15:49:34 <lanius> xorAxAx: you mean a virtual root item which represents the global edit-log
2007-07-23T15:49:46 <xorAxAx> well, not a virtual root item .. anything virtual
2007-07-23T15:49:54 <xorAxAx> i dislike this b0rked root item stuff :)
2007-07-23T15:49:59 <lanius> but some item which is not a page
2007-07-23T15:50:04 <xorAxAx> yes
2007-07-23T15:50:04 <lanius> no data in it
2007-07-23T15:50:06 <xorAxAx> yes
2007-07-23T15:50:39 <lanius> well two problems with that: 1.) it sounds like putting something onto something where it does not really belong 2.) you still have to hold the data on two places, on the global and on the local place
2007-07-23T15:50:59 <xorAxAx> yes, redundancies
2007-07-23T15:51:07 <xorAxAx> you could also provide a kind of view
2007-07-23T15:51:15 <xorAxAx> that exposes a sequence of revisions to the caller
2007-07-23T15:51:26 <xorAxAx> and internally, its just links to the items/rev nos
2007-07-23T15:52:02 <xorAxAx> but thats much slower than the old linear file because of the indirections and the parsing overhead
2007-07-23T15:52:09 <lanius> yes but i think the base view would be the global edit-log from which you could derive the local edit-log
2007-07-23T15:52:20 <xorAxAx> the local edit-log is a list of revisions
2007-07-23T15:52:34 <xorAxAx> thats perfectly fine for performance and complexity reasons
2007-07-23T15:52:37 <lanius> right
2007-07-23T15:53:00 <lanius> but when you want to construct a global edit-log out of the local edit-log you would have to parse ALL items for their changes
2007-07-23T15:53:24 <xorAxAx> yeah, but you dont need to, it amends the list when creating a revision of course
2007-07-23T15:53:34 <lanius> which would be a cache then
2007-07-23T15:53:41 <xorAxAx> well
2007-07-23T15:53:42 <xorAxAx> no
2007-07-23T15:53:54 <xorAxAx> a list of pointers is less than a cache
2007-07-23T15:53:59 <xorAxAx> it only contains refs, no data
2007-07-23T15:54:13 <lanius> but somehow it must be saved in the cache
2007-07-23T15:54:31 <xorAxAx> yes, or its hidden in the backend if that makes sense
2007-07-23T16:03:07 <lanius> sounds good
2007-07-23T16:03:18 <xorAxAx> maybe draft something in the wiki
2007-07-23T16:05:20 <lanius> i don't think there is much to draft
2007-07-23T16:05:38 <lanius> the thing is more about Thomas saying yes and me implementing it ,)
2007-07-23T16:05:40 <xorAxAx> well, then dont introduce layering violations :)
2007-07-23T16:07:32 <lanius> the caching module is just another module which can be used everywhere, isn't it?
2007-07-23T16:07:41 <xorAxAx> you want to use the caching module?
2007-07-23T16:07:45 <lanius> why not
2007-07-23T16:07:49 <xorAxAx> that will slow it down even more
2007-07-23T16:08:01 <xorAxAx> i was thinking about a list on the backend level
2007-07-23T16:08:14 <lanius> stored in cache or on disk=
2007-07-23T16:08:21 <xorAxAx> in the backend :)
2007-07-23T16:08:28 <xorAxAx> but maybe thats a stupid idea :)
2007-07-23T16:08:36 <lanius> maybe not
2007-07-23T16:09:34 * ThomasWaldmann being back after doing some orders
2007-07-23T16:09:38 <lanius> :D
2007-07-23T16:19:50 <lanius> ThomasWaldmann: so any comments
2007-07-23T16:21:31 <dreimark> ThomasWaldmann: http://moinmoin.wikiwikiweb.de/MoinMoinBugs/1.7devHeadersAlreadySentException
2007-07-23T16:37:28 <dreimark> strange that we get it only in that case. the other if requests before seemed not to be used ?
2007-07-23T17:23:16 <dreimark> bbl
2007-07-23T17:37:41 <ThomasWaldmann> re
2007-07-23T17:37:54 <ThomasWaldmann> sorry, had a rather long customer talk
2007-07-23T17:39:32 * ThomasWaldmann loves descriptive commit comments like 'fixes around'
2007-07-23T17:39:55 <grzywacz> Hm. Sometimes too descriptive is bad as well.
2007-07-23T17:39:56 <grzywacz> ;]
2007-07-23T17:40:51 <ThomasWaldmann> "misc. changes" :-P
2007-07-23T17:42:50 <ThomasWaldmann> lanius: did you do any concurrency tests?
2007-07-23T17:46:34 <ThomasWaldmann> lanius: how does refresh() work for multiple concurrent processes?
2007-07-23T17:47:20 <ThomasWaldmann> looks like it only works for same process and the other will die dumb :)
2007-07-23T17:48:12 <vpv> omg, this is frustrating, if I make docbook xml from texinfo files with makeinfo --docbook, the resulting data crashes python's xmlrpc handler :@
2007-07-23T17:48:34 <ThomasWaldmann> (a multi-process event system would be nice, but we don't have that yet. events are also only "same process")
2007-07-23T17:49:03 <ThomasWaldmann> vpv: remember the docbook branch :)
2007-07-23T17:49:05 <xorAxAx> vpv: ?
2007-07-23T17:49:24 <xorAxAx> vpv: how do you pass them?
2007-07-23T17:49:46 <ThomasWaldmann> i guess verbatim :-P
2007-07-23T17:49:52 <grzywacz> lol
2007-07-23T17:50:08 <vpv> well, yes, any ideas?
2007-07-23T17:50:35 <ThomasWaldmann> either escape/unescape or use binary
2007-07-23T17:50:52 <grzywacz> vpv, it should be base64-encoded, no?
2007-07-23T17:52:37 <ThomasWaldmann> grzywacz: in case of you feel bored, how about a multi process capable event system? :D
2007-07-23T17:53:20 <vpv> so if I encode the data, will xmlrpclib decode it back when making the page add call?
2007-07-23T17:54:02 <ThomasWaldmann> (or at least adding a big hint: Do NOT use this if more than the same process needs to be notified.)
2007-07-23T17:54:46 <ThomasWaldmann> vpv: look into wikirpc.py
2007-07-23T17:55:45 <ThomasWaldmann> (esp. the v1 part - it had to encode/decode quite a lot because text was defined as ASCII-only, so you even had to treat utf-8 like binary)
2007-07-23T17:56:37 <ThomasWaldmann> http://hg.moinmo.in/moin/1.7-storage-hwendel/rev/ac106da04e3c I guess that won't work correctly for multiple parallel long-running moins (e.g. fastcgi)
2007-07-23T17:56:59 <vpv> ThomasWaldmann: seems I can't find that in 1.7...
2007-07-23T17:57:14 <ThomasWaldmann> sorry, xmlrpc/__init__.py
2007-07-23T17:57:48 <ThomasWaldmann> wiki* -> */__init__
2007-07-23T17:58:04 <ThomasWaldmann> except for wikiutil, too much work
2007-07-23T17:58:12 <vpv> ok
2007-07-23T17:59:00 <grzywacz> ThomasWaldmann, um, elaborate please ;-p
2007-07-23T18:00:00 <ThomasWaldmann> imagine 2 moin processes having some sort of in-process cache
2007-07-23T18:00:13 <grzywacz> What for?
2007-07-23T18:00:26 <ThomasWaldmann> then you delete a page by using process 1 (cache for page metadata for example)
2007-07-23T18:00:53 <ThomasWaldmann> it then sends an even PageDeleted(somepage) and updates its own cache by handling it
2007-07-23T18:01:06 <ThomasWaldmann> process 2 won't notice it
2007-07-23T18:01:19 <ThomasWaldmann> (and live on with wrong cache content)
2007-07-23T18:01:55 <ThomasWaldmann> we already talked about that 1m ago :)
2007-07-23T18:09:02 <ThomasWaldmann> grzywacz: currently, global edit-log is used as kind of inter-process page event notification, by just tracking filepos / filelength
2007-07-23T18:09:52 <xorAxAx> well, itemcache was dropped
2007-07-23T18:09:59 <xorAxAx> in lanius' repo
2007-07-23T18:11:35 <ThomasWaldmann> xorAxAx: http://hg.moinmo.in/moin/1.7-storage-hwendel/rev/ac106da04e3c see external.py
2007-07-23T18:12:24 <ThomasWaldmann> (and below also storagenotify.py)
2007-07-23T18:12:53 <xorAxAx> well, thats certainly a differnetly shaped cache
2007-07-23T18:13:32 <ThomasWaldmann> yeah, mine was square, this is triangular :D
2007-07-23T18:14:38 <xorAxAx> maybe lanius can give us an overview over the cached stuff, there are too many underscores in the patch to allow me to get an overview :)
2007-07-23T18:17:27 <ThomasWaldmann> + # TODO: fix this
2007-07-23T18:17:39 <ThomasWaldmann> fix what? X)
2007-07-23T19:33:25 <johill> I had a bunch of new ideas :)
2007-07-23T19:38:57 <ThomasWaldmann> tell us :)
2007-07-23T19:39:03 <johill> :)
2007-07-23T19:39:12 <johill> well for example
2007-07-23T19:39:16 <johill> I had allowed **kw
2007-07-23T19:39:20 <johill> and _non_ascii_kwargs
2007-07-23T19:39:30 <johill> but decided that just _kwargs as a special parameter that gets a dict is much nicer
2007-07-23T19:39:40 <johill> also; I cleaned up everything and allowed dreimark's choice thing
2007-07-23T19:40:00 <johill> and got all errors handled by checking myself instead of getting TypeError
2007-07-23T19:40:25 <johill> oh and that also cleans up the Macro() vs Macro(,) thing
2007-07-23T19:40:32 <johill> (gets None in all cases)
2007-07-23T19:41:23 <ThomasWaldmann> i dont quite understand the kw stuff (except that currently you can't have unicode names in fn header)
2007-07-23T19:41:37 <ThomasWaldmann> (wait for python3000 :)
2007-07-23T19:41:38 <johill> yeah that's the thing
2007-07-23T19:41:50 <johill> and you can't have unicode names passed into **kw
2007-07-23T19:42:03 <johill> so I introduced the special _non_ascii_kwargs parameter that gets those
2007-07-23T19:42:30 <johill> which I just decided was stupid because unifying both is nicer and I removed the special one and added _kwargs as a special one that gets all that haven't been taken
2007-07-23T19:42:38 <johill> if you don't have _kwargs then it's still invalid
2007-07-23T19:44:07 <ThomasWaldmann> and if you want to use keyword args, you have to write _kwargs now?
2007-07-23T19:44:17 <johill> yeah
2007-07-23T19:44:29 <ThomasWaldmann> kw is shorter
2007-07-23T19:44:48 <johill> right, but you might rightfully have a macro that takes kw as a parameter
2007-07-23T19:45:22 <johill> not many macros are going to use that anyway
2007-07-23T19:45:28 <johill> I guess the action macro would, but that's probably about it
2007-07-23T19:45:30 <ThomasWaldmann> ok
2007-07-23T19:46:02 <johill> and the last idea was to introduce a wrapper class required_arg
2007-07-23T19:46:04 <johill> so you can write
2007-07-23T19:46:14 <johill> def macro_X(macro, i=required_arg(int)):
2007-07-23T19:46:20 <johill> and then the generic code checks that it was given
2007-07-23T19:48:28 <ThomasWaldmann> sounds ok
2007-07-23T19:49:02 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2538:070294763d3e 1.7/MoinMoin/ (macro/__init__.py wikiutil.py): clean up parser argument handling, allow tuples for choices
2007-07-23T19:49:03 <johill> spam spam spam :)
2007-07-23T19:49:04 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2539:9dd4a7a164e4 1.7/MoinMoin/macro/Hits.py: use new features in Hits macro
2007-07-23T19:49:05 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2540:fa4252cb4077 1.7/MoinMoin/ (macro/__init__.py wikiutil.py): move macro invoker to wikiutil
2007-07-23T19:49:13 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2541:5477d7f5ad64 1.7/MoinMoin/macro/__init__.py: small try/except cleanup in macro code
2007-07-23T19:49:16 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2542:07ac71a1d9d2 1.7/MoinMoin/wikiutil.py: fix bug with positional args and the special ones
2007-07-23T19:49:17 * ThomasWaldmann drives home :)
2007-07-23T19:49:24 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2543:116ab80df98b 1.7/MoinMoin/_tests/test_wikiutil.py: tests for the extension function invoker
2007-07-23T19:49:30 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2544:bc49e37d2649 1.7/MoinMoin/_tests/test_wikiutil.py: quoting tests for the parameter parser
2007-07-23T19:49:38 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2545:3ef21e54f940 1.7/MoinMoin/wikiutil.py: fix quoting of quote at end of item in parameter parser
2007-07-23T19:49:46 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2546:56dbfbde79fe 1.7/MoinMoin/ (_tests/test_wikiutil.py macro/__init__.py wikiutil.py): better error checking in parameter parser, no more TypeError
2007-07-23T19:49:52 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2547:c2ef5800253f 1.7/MoinMoin/ (_tests/test_wikiutil.py wikiutil.py): consolidate extra keyword args handling, no more **kw/_non_ascii_kw but rather just _kwargs
2007-07-23T19:50:01 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2548:5db66a25f0ba 1.7/MoinMoin/ (_tests/test_wikiutil.py wikiutil.py): introduce required_arg wrapper class
2007-07-23T19:51:07 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2549:fb88a3706761 1.7/MoinMoin/macro/__init__.py: macro invoker doesn't need to catch TypeError any more
2007-07-23T20:14:40 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2550:2523642e9516 1.7/MoinMoin/ (_tests/test_wikiutil.py wikiutil.py): put extra additional arguments into _trailing_args in all cases
2007-07-23T20:15:47 * xorAxAx likes the granularity :)
2007-07-23T20:17:23 <johill> hm?
2007-07-23T20:18:51 <johill> ok I'm much happier with it now :)
2007-07-23T20:21:16 <xorAxAx> johill: compared to others, you split up the commits well :)
2007-07-23T20:22:14 <johill> oh
2007-07-23T20:22:20 <johill> you learn that when you hack the kernel ;)
2007-07-23T20:23:33 <xorAxAx> hehe
2007-07-23T20:23:56 <johill> I have a patch that makes the parameter parser able to use different key signs
2007-07-23T20:24:12 <johill> and return a tuple-list [(key,value), (key,value)] instead of a dict
2007-07-23T20:24:21 <johill> I thought that could be useful for parsing searches
2007-07-23T20:24:41 <johill> so it can be like title:"a b c" title:"d e f" "x y"
2007-07-23T20:24:54 <johill> to require both 'a b c' and 'd e f' in the title and 'x y' in the body
2007-07-23T20:24:59 <johill> but I'm not sure if it's even possible to hook that in
2007-07-23T20:25:30 <johill> think I should push it anyway?
2007-07-23T20:26:06 <xorAxAx> not sure
2007-07-23T20:27:00 <johill> me neither really :)
2007-07-23T20:27:08 <johill> was just looking for places that could use it too
2007-07-23T20:27:24 <xorAxAx> the search has a complex modifier syntax
2007-07-23T20:27:36 <johill> like?
2007-07-23T20:28:13 <johill> I mean, why does that matter?
2007-07-23T20:28:24 <xorAxAx> no idea :)
2007-07-23T20:28:38 <xorAxAx> its parsing at least
2007-07-23T20:28:53 <johill> basically the parser would give you a list of things the user entered
2007-07-23T20:29:09 <johill> [u'abc', u'-def', u'x y z']
2007-07-23T20:29:36 <johill> and then maybe [(u'title', u'jj'), (u'title': u'dd')]
2007-07-23T20:29:59 <johill> the advantage would just be that all the quote parsing is there and well-tested
2007-07-23T20:34:49 <johill> xorAxAx: oh btw, I got rid of all the internal errors too by that rewrite that checks all the args
2007-07-23T20:36:23 <xorAxAx> ah
2007-07-23T20:37:44 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2551:f3b684afca81 1.7/MoinMoin/_tests/test_wikiutil.py: make sourcecode checker happy on wikiutil tests
2007-07-23T20:41:42 <johill> uh
2007-07-23T20:41:48 <johill> when I install python-crack the test fails
2007-07-23T20:41:51 <johill> > IOError: [Errno 2] No such file or directory: '/var/cache/cracklib/cracklib_dict.hwm'
2007-07-23T20:42:25 <ThomasWaldmann> you need to install some wordlists (wenglish) and run the cron script to make those dicts
2007-07-23T20:42:35 <johill> yeah but the test should be skipped :)
2007-07-23T20:43:04 <ThomasWaldmann> there is some attempt to do that, it seems to need fixing
2007-07-23T20:44:04 <johill> ah heh
2007-07-23T20:44:11 <johill> the problem is the self.request.cfg.password_checker
2007-07-23T20:44:18 <johill> it doesn't have the workaround
2007-07-23T20:44:33 <johill> but why do you use that? seems that should be a separate test
2007-07-23T20:44:51 <xorAxAx> hmm?
2007-07-23T20:44:56 <xorAxAx> "separate test"?
2007-07-23T20:45:00 <johill> yeah
2007-07-23T20:45:01 <johill> see
2007-07-23T20:45:11 <johill> the test here checks whether self.request.cfg.password_checker is set
2007-07-23T20:45:17 <johill> and otherwise skips the test
2007-07-23T20:45:20 <johill> don't understand
2007-07-23T20:45:28 <johill> shouldn't it test the default checker?
2007-07-23T20:45:48 <johill> or if it's not default, the shipped checker?
2007-07-23T20:48:12 <xorAxAx> ah
2007-07-23T20:48:15 <xorAxAx> makes sense
2007-07-23T20:48:44 <johill> just seemed a bit weird to test something from the config
2007-07-23T20:49:30 <ThomasWaldmann> the shipped checker is in the config
2007-07-23T20:50:10 <johill> but then why check if it is assigned?
2007-07-23T20:50:58 <johill> guess it doesn't really make a difference
2007-07-23T20:55:31 <ThomasWaldmann> you COULD disable it, e.g. by assigning None, then there is nothing to test any more
2007-07-23T20:55:51 <ThomasWaldmann> try:
2007-07-23T20:55:52 <ThomasWaldmann> crack.FascistCheck("a12fv./ZX47") # this should not raise an exception
2007-07-23T20:55:55 <ThomasWaldmann> except:
2007-07-23T20:55:57 <ThomasWaldmann> py.test.skip("python-crack is not working correctly (did you forget to build the dicts?)")
2007-07-23T20:56:08 <ThomasWaldmann> the question is rather why this didnt work for you
2007-07-23T20:59:14 <vpv> rayvd: you're not on #fedora-docs? paulobanon has agreed to sponsor our moin RFRs
2007-07-23T21:18:10 <rayvd> great vpv... sorry i have been afk with work stuff today :-/
2007-07-23T21:18:12 <rayvd> i will be back shortly.
2007-07-23T21:22:03 <neagulm> moin dreimark
2007-07-23T21:22:31 <vpv> rayvd: no worries
2007-07-23T21:22:33 <xorAxAx> neagulm: hi, didnt you want to phone me? :)
2007-07-23T21:22:40 <neagulm> dreimark, finally I am out of my intellectual deadlock
2007-07-23T21:23:00 <neagulm> xorAxAx, dreimark i had to do some thinking about the n-way classification
2007-07-23T21:24:05 <neagulm> xorAxAx, dreimark with help from one teacher from school I think I found a solution that my apply good for MoinMoin
2007-07-23T21:24:15 <xorAxAx> ok
2007-07-23T21:25:22 <neagulm> xorAxAx, dreimark prof. Zaharie has suggested to use a incremental clustering algorithm
2007-07-23T21:26:38 <neagulm> xorAxAx, dreimark she suggested to use an "ART neural network" for clustering and that I will be relatively easy to implement
2007-07-23T21:28:05 <xorAxAx> sounds good
2007-07-23T21:29:27 <dreimark_> neagulm: ok, you are telling about the recommendation (I missed the beginning)
2007-07-23T21:30:55 <dreimark_> johill: marvelous :)
2007-07-23T21:31:38 <neagulm> dreimark_, let's say wiki page grouping, categorization or clustering. This should be the base of the recommendation system or anything we want to implement over it
2007-07-23T21:31:44 <xorAxAx> dreimark_: ThomasWaldmann said that current tip is currently broken because of a change of yours
2007-07-23T21:31:52 <xorAxAx> dreimark_: http://paste.pocoo.org/show/1922
2007-07-23T21:34:25 <dreimark_> not good
2007-07-23T21:37:05 <dreimark_> neagulm: fine,
2007-07-23T21:42:23 <dreimark_> xorAxAx: ThomasWaldmann do you know how it was called http://en.wikipedia.org/wiki/Special:Search?go=Go&search=Action(....)
2007-07-23T21:42:31 <dreimark_> xorAxAx: ThomasWaldmann do you know how it was called [ [Action(....)] ]
2007-07-23T21:43:03 <ThomasWaldmann> action(edit) maybe?
2007-07-23T21:43:08 <dreimark_> this works
2007-07-23T21:43:29 <xorAxAx> so its his own code?
2007-07-23T21:43:36 <ThomasWaldmann> the problem is that the split can give 1 or 2 items as result
2007-07-23T21:44:15 <ThomasWaldmann> so you either have to "fill it up", or assign to one item on the left and then check its length
2007-07-23T21:44:53 <dreimark_> ThomasWaldmann: you were right, i was not uptodate here
2007-07-23T21:48:29 <ThomasWaldmann> i sometimes used the fillup method
2007-07-23T21:49:31 <ThomasWaldmann> a,b,c,d = (x.split(',') + [None] * 3)[:4]
2007-07-23T21:51:58 <grzywacz> ...
2007-07-23T21:52:07 <xorAxAx> grzywacz: my student!
2007-07-23T21:52:21 <grzywacz> xorAxAx, yes my hmre?!
2007-07-23T21:52:32 <xorAxAx> grzywacz: you printed three dots. why?
2007-07-23T21:53:08 <grzywacz> xorAxAx, in response to Thomas' crazy line!
2007-07-23T21:53:40 <xorAxAx> well
2007-07-23T21:53:49 <xorAxAx> thats rsplit for legacy guys
2007-07-23T21:54:00 <xorAxAx> hmm, no
2007-07-23T21:55:56 <CIA-27> moin: Karol 'grzywacz' Nowak <grzywacz@sul.uni.lodz.pl> * 2507:a791fc964e09 1.7-jabber-knowak/jabberbot/ (commands.py xmlrpcbot.py xmppbot.py): Attempt to move i18n from xmlrpc to xmpp thread.
2007-07-23T21:55:59 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2508:ea255685d6b0 1.7-jabber-knowak/MoinMoin/ (_tests/test_wikiutil.py macro/__init__.py wikiutil.py): add macro argument parser and use it to invoke macros with args directly
2007-07-23T21:56:06 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2509:552e748b86cd 1.7-jabber-knowak/MoinMoin/wikiutil.py: add get_{boolean,int,float,unicode} helpers
2007-07-23T21:56:17 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2510:a55a20b092a2 1.7-jabber-knowak/MoinMoin/wikiutil.py: make get_int safer, fix get_unicode(arg=u'')
2007-07-23T21:56:25 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2511:f6a777f45559 1.7-jabber-knowak/MoinMoin/ (_tests/test_wikiutil.py wikiutil.py): clarify None behaviour for keys, add tests
2007-07-23T21:56:37 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2512:e85e1ebb3459 1.7-jabber-knowak/MoinMoin/wikiutil.py: remove bogus _ variable from get_unicode
2007-07-23T21:56:45 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2513:68496057ee03 1.7-jabber-knowak/MoinMoin/ (_tests/test_wikiutil.py wikiutil.py): wikiutil.get_* tests + fixes
2007-07-23T21:56:58 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2514:6de0e14591f7 1.7-jabber-knowak/MoinMoin/ (macro/MonthCalendar.py macro/__init__.py wikiutil.py): macro.format_error to format what a macro shows in case of ValueError/TypeError happening, refactor MonthCalendar to use new arg parsing
2007-07-23T21:57:07 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2515:9065adc695ca 1.7-jabber-knowak/MoinMoin/macro/Hits.py: refactored Hits macro for new arg parser
2007-07-23T21:57:15 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2516:41bc022c2160 1.7-jabber-knowak/MoinMoin/macro/__init__.py: use new arg parser for buitlin macros
2007-07-23T21:57:24 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2517:05c9d52dcf5d 1.7-jabber-knowak/MoinMoin/parser/ (_tests/test_text_moin_wiki.py text_moin_wiki.py): text_moin_wiki:fixed wrong pre section, skipped a test because not implemented yet
2007-07-23T21:57:35 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2518:50b12f981890 1.7-jabber-knowak/MoinMoin/macro/__init__.py: builtin macros: fix bug, handle errors giving help to users
2007-07-23T21:57:43 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2519:fc5b843af8f1 1.7-jabber-knowak/MoinMoin/parser/ (_tests/test_text_moin_wiki.py text_moin_wiki.py): merge main
2007-07-23T21:57:51 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2520:5f299f9ba77c 1.7-jabber-knowak/MoinMoin/macro/__init__.py: externalize InterWiki macro, refactor it to use the formatter
2007-07-23T21:57:58 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2521:02766a483465 1.7-jabber-knowak/MoinMoin/macro/__init__.py: builtin macros: just reorder, no other changes
2007-07-23T21:58:05 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2522:b4118a77105d 1.7-jabber-knowak/MoinMoin/macro/ (__init__.py PageCount.py): externalize PageCount macro
2007-07-23T21:58:10 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2523:d73957833b45 1.7-jabber-knowak/MoinMoin/macro/InterWiki.py: add InterWiki.py (see cs 5f299f9ba77c)
2007-07-23T21:58:17 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2524:ee8d1a2cc252 1.7-jabber-knowak/MoinMoin/macro/__init__.py: XSS fix for format_error
2007-07-23T21:58:24 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2525:35dcbfc32a57 1.7-jabber-knowak/MoinMoin/macro/_tests/test_Hits.py: Fix Hits macro tests
2007-07-23T21:58:31 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2526:ce7fc6ba9b3a 1.7-jabber-knowak/MoinMoin/_tests/test_sourcecode.py: test_sourcecode: only test recently touched files
2007-07-23T21:58:36 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2527:860e7bf6c4cc 1.7-jabber-knowak/MoinMoin/parser/_tests/test_text_moin_wiki.py: add failing {{{ }}} test
2007-07-23T21:58:46 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2528:4f6f7e56eb40 1.7-jabber-knowak/MoinMoin/wikiutil.py: fix error message for get_* (arg parser)
2007-07-23T21:58:54 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2529:1ae621d83c5b 1.7-jabber-knowak/MoinMoin/ (_tests/test_wikiutil.py wikiutil.py): fix arg parser bug, test for this bug
2007-07-23T21:59:00 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2530:df375ab9fbde 1.7-jabber-knowak/MoinMoin/macro/__init__.py: automatically convert macro args with defaults to the default's type
2007-07-23T21:59:10 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2531:0a8fc701e40d 1.7-jabber-knowak/MoinMoin/macro/__init__.py: use the new automatic macro arg conversion in some internal macros
2007-07-23T21:59:18 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2532:1184922dc419 1.7-jabber-knowak/MoinMoin/macro/__init__.py: some more explanations of the type magic for macro args
2007-07-23T21:59:28 <CIA-27> moin: Johannes Berg <johannes AT sipsolutions DOT net> * 2533:cc125871c8a0 1.7-jabber-knowak/MoinMoin/macro/__init__.py: whoops, get_boolean was renamed to get_bool
2007-07-23T21:59:34 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2534:19ba0aa5ccdf 1.7-jabber-knowak/MoinMoin/parser/_tests/test_text_moin_wiki.py: test_text_moin_wiki:fixed output from testMultipleShortPreSections
2007-07-23T21:59:41 <CIA-27> (19 lines omitted)
2007-07-23T22:03:01 <lanius> ThomasWaldmann: i just realized that
2007-07-23T22:03:05 <lanius> ThomasWaldmann: sadly
2007-07-23T22:03:49 <ThomasWaldmann> yeah, it was so pretty :)
2007-07-23T22:13:45 <ThomasWaldmann> xorAxAx: can you please open the editor on http://test17.wikiwikiweb.de/TestWiki17MainBranch ?
2007-07-23T22:14:44 <xorAxAx> ThomasWaldmann: which browser?
2007-07-23T22:14:50 <ThomasWaldmann> doesnt matter
2007-07-23T22:15:15 <xorAxAx> works using the text editor
2007-07-23T22:15:33 <ThomasWaldmann> ok, reproduced that bug
2007-07-23T22:15:36 <ThomasWaldmann> close it now
2007-07-23T22:15:54 <xorAxAx> i already edited
2007-07-23T22:16:04 <ThomasWaldmann> open it again
2007-07-23T22:16:36 <dreimark_> once again Formatting a text that is being formatted?!
2007-07-23T22:16:52 <ThomasWaldmann> xorAxAx: it's open?
2007-07-23T22:17:03 <xorAxAx> now it is
2007-07-23T22:17:07 <xorAxAx> but why do you need me?
2007-07-23T22:17:08 <ThomasWaldmann> dreimark_: no, the headers twice bug
2007-07-23T22:17:14 <xorAxAx> i am seeing a header bug
2007-07-23T22:17:15 <grzywacz> emit_http_headers called multiple (2) times! Headers: ['Content-Type: text/html; charset=utf-8', 'Status: 200 OK']
2007-07-23T22:17:19 <grzywacz> Same here.
2007-07-23T22:17:34 <ThomasWaldmann> grzywacz: where?
2007-07-23T22:17:53 <xorAxAx> Page.py:1096
2007-07-23T22:17:56 <xorAxAx> LOL, the counter is funny
2007-07-23T22:18:00 <grzywacz> ThomasWaldmann, on the url you've just provided.
2007-07-23T22:18:02 <ThomasWaldmann> oops, I thought I had fixed it
2007-07-23T22:18:13 <xorAxAx> as if the code allowed for more than 2 times of that exception :)
2007-07-23T22:18:45 <grzywacz> xorAxAx, you never know what breakage is sneaking upon you!
2007-07-23T22:18:47 <grzywacz> ;)
2007-07-23T22:18:59 <xorAxAx> yeah
2007-07-23T22:19:30 <ThomasWaldmann> xorAxAx: the exception is newer than the counter
2007-07-23T22:19:46 <xorAxAx> yes, printing the counter is the funny part of that
2007-07-23T22:19:54 <ThomasWaldmann> yeah
2007-07-23T22:24:08 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2119:7424ea30a256 1.6/MoinMoin/macro/Action.py: macro.Action:correct splitting of action and args (filling up)
2007-07-23T22:24:18 <dreimark_> very clever ThomasWaldmann
2007-07-23T22:25:52 <ThomasWaldmann> dreimark_: what, the exception counter? :D
2007-07-23T22:26:33 <ThomasWaldmann> I guess I am not in the right mood for PageEditor today X)
2007-07-23T22:26:34 <dreimark_> the filling up trick
2007-07-23T22:26:58 <dreimark_> + [None] * 2
2007-07-23T22:35:18 <dreimark_> hg status
2007-07-23T22:35:26 <dreimark_> hehe, wrong window
2007-07-23T22:36:36 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2552:88ea1324d55e 1.7/MoinMoin/macro/Action.py: macro.Action.py: fixed bug in in using Parameters for e.g. Edit (thanks to Johannes Berg)
2007-07-23T22:36:37 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> * 2553:eb3297891475 1.7/MoinMoin/macro/Action.py: macro.Action:correct splitting of action and args (filling up)
2007-07-23T22:39:09 <dreimark_> ThomasWaldmann: that HeadersAlreadySentException bug needs to be checked for 1.6 too (I guess its there too)
2007-07-23T22:39:28 <ThomasWaldmann> sure
2007-07-23T22:46:22 <vpv> anyone care to elaborate on what I should do with the xmlrpc page data encoding/decoding? should I subclass XmlRpcBase and make my own _instr, _outstr or what?
2007-07-23T22:46:42 <xorAxAx> no, you shouldnt
2007-07-23T22:46:45 <xorAxAx> whats the issue?
2007-07-23T22:47:54 <vpv> when I have XML data in an xmlrpc multicall, the xmlrpc handling may crash because if it's not well-formed
2007-07-23T22:48:58 <vpv> so as it was discussed earlier, I should probably encode the XML "payload" data instead of sending it verbatim, but how?
2007-07-23T22:50:30 <xorAxAx> vpv: first find out whats incorrectly encoded
2007-07-23T22:50:40 <xorAxAx> doesnt it escape the message?
2007-07-23T22:53:53 <vpv> this is what it says, http://paste.pocoo.org/show/1924
2007-07-23T22:54:32 <vpv> so there's something in the XML data I'm trying to add to a page that messes up Python's XML parser
2007-07-23T22:56:24 <xorAxAx> well, look at the data that it send via xmlrpc
2007-07-23T22:56:27 <xorAxAx> and spot the problem
2007-07-23T23:11:27 <vpv> I haven't been able to find the exact place, but I'm thinking it might have problems with names like these: Gaël Quéri
2007-07-23T23:11:47 <vpv> Because those characters are used a lot in the area where it crashes
2007-07-23T23:12:17 <xorAxAx> sigh, you just need to look at the generated body
2007-07-23T23:12:22 <xorAxAx> ok, then you have an encoding problem
2007-07-23T23:12:28 <xorAxAx> do you use xmlrpc2?
2007-07-23T23:12:32 <vpv> yes
2007-07-23T23:13:10 <xorAxAx> whats the type of the object you use on the caller side?
2007-07-23T23:14:14 <xorAxAx> vpv:
2007-07-23T23:14:38 <vpv> I'm not sure I understood the question, but it's a multicall :/
2007-07-23T23:15:28 <xorAxAx> sigh
2007-07-23T23:15:31 <xorAxAx> point me to the code
2007-07-23T23:17:09 <vpv> http://hg.moinmo.in/moin/1.7-maninfo-vpv/file/da96ec31ad6a/MoinMoin/script/xmlrpc/manimport.py
2007-07-23T23:17:26 <xorAxAx> line?
2007-07-23T23:17:28 <vpv> the multicall is run on line 785
2007-07-23T23:18:24 <xorAxAx> vpv: well, which type have the objects pagename and lines?
2007-07-23T23:18:40 <xorAxAx> they need to be unicode
2007-07-23T23:19:17 <vpv> ok
2007-07-23T23:20:14 <vpv> that does help a bit, thanks, I tried actually to convert the "lines" into unicode, but then it raises errors in about the same lines of the XML
2007-07-23T23:20:23 <xorAxAx> ThomasWaldmann: is CIA set up for the docbook branch?
2007-07-23T23:20:46 <xorAxAx> vpv: a bit only? is the error gone now?
2007-07-23T23:21:49 <vpv> no, but I didn't know the args should generally be unicode, I'll do a test now...
2007-07-23T23:22:25 <xorAxAx> they shouldnt be generally
2007-07-23T23:23:24 <vpv> ok, so now I get UnicodeDecodeError: 'ascii' codec can't decode byte 0xe9 in position 14: ordinal not in range(128)
2007-07-23T23:24:08 <vpv> if I use unicode() in line 760 (and of course the start of the string is made with u"...")
2007-07-23T23:25:13 <xorAxAx> you shouldnt use unicode()
2007-07-23T23:25:16 <xorAxAx> but str.decode
2007-07-23T23:26:27 <xorAxAx> dreimark_: ping
2007-07-23T23:26:38 <dreimark_> xorAxAx: pong
2007-07-23T23:26:41 <vpv> ok, I'll try that after I've read the discussion on #moin, which seem interesting
2007-07-23T23:26:48 <xorAxAx> dreimark_: you arent into XML, are you? :)
2007-07-23T23:26:56 <xorAxAx> vpv: hehe
2007-07-23T23:27:00 <dreimark_> no
2007-07-23T23:27:12 <dreimark_> may be he can setup http://moinmoin.wikiwikiweb.de/MoinBounties
2007-07-23T23:41:57 <ThomasWaldmann> xorAxAx: cia should work for db branch
2007-07-23T23:43:24 <xorAxAx> ok
2007-07-23T23:49:26 <vpv> I got it to work when I set str.decode("utf-8", "ignore"), but Moin won't display the page correctly, because there's a mismatched tag
2007-07-23T23:49:58 <xorAxAx> then utf-8 is not the right encoding, vpv
2007-07-23T23:54:04 <vpv> even if I set it to "ascii" it won't help, I think the point is that the source document is not valid
2007-07-23T23:54:36 <xorAxAx> vpv: ascii is the strongest choice - it means that the document doesnt contain any umlauts etc.
2007-07-23T23:54:44 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> * 2203:100ffdaceedb 1.6-docbook-mvirkill/MoinMoin/ (formatter/text_docbook.py parser/text_docbook.py): more PEP8 cleanup
2007-07-23T23:54:48 <xorAxAx> vpv: you need to know the correct encoding at this point to do anything sane
2007-07-23T23:54:56 <ThomasWaldmann> works :)
2007-07-23T23:56:04 <vpv> xorAxAx: ok, but what I mean is that I think it's not valid XML, it might be just tagsoup
2007-07-23T23:58:52 <xorAxAx> vpv: thats unrelated at this point
2007-07-23T23:58:58 <xorAxAx> when it comes to encodings MoinMoin: MoinMoinChat/Logs/moin-dev/2007-07-23 (last edited 2007-10-29 19:10:50 by localhost)
