2007-09-23T00:51:56 <ThomasWaldmann> wtf - if i give FCKeditor an img with alt attr, it displays the alt text instead of the image!?
2007-09-23T00:54:37 <TheSheep> ThomasWaldmann: maybe the url is wrong?
2007-09-23T00:58:09 <ThomasWaldmann> forget it. retried and it works now
2007-09-23T01:00:37 <ThomasWaldmann> hmm, i need some kind of "userdef" attribute for html tags to roundtrip some stuff.
2007-09-23T01:01:04 <ThomasWaldmann> currently it uses title attr for that, but that spoils real use of title attr
2007-09-23T01:01:33 <ThomasWaldmann> e.g. it puts "attachment:foo.txt" into title of attachment links to easier detect them
2007-09-23T01:05:50 <TheSheep> ThomasWaldmann: so you need a throwaway attribute that would still validate?
2007-09-23T01:06:03 <ThomasWaldmann> yeah, that would be best
2007-09-23T01:06:09 <TheSheep> ThomasWaldmann: how about just inserting comments?
2007-09-23T01:06:28 <TheSheep> <a ...><!-- whatver you need -->...</a>
2007-09-23T01:06:36 <ThomasWaldmann> that no attr
2007-09-23T01:06:59 <ThomasWaldmann> (attr access would be easiest)
2007-09-23T01:07:22 <TheSheep> well, I don't think there is a scratchit attribute without any side effects
2007-09-23T01:07:31 <ThomasWaldmann> something like a nop="dont care"
2007-09-23T01:08:12 <TheSheep> ThomasWaldmann: name?
2007-09-23T01:08:31 <TheSheep> ThomasWaldmann: rel?
2007-09-23T01:08:43 <ThomasWaldmann> dupe names allowed?
2007-09-23T01:08:48 <TheSheep> nope
2007-09-23T01:08:57 <TheSheep> ThomasWaldmann: you need it on img or on a?
2007-09-23T01:09:05 <ThomasWaldmann> both
2007-09-23T01:09:15 <TheSheep> class? :)
2007-09-23T01:09:19 <ThomasWaldmann> (and maybe even other?)
2007-09-23T01:09:28 <ThomasWaldmann> class is needed for display
2007-09-23T01:09:52 <TheSheep> sure, but you can always add something at the end
2007-09-23T01:09:56 <TheSheep> after a space
2007-09-23T01:10:05 <ThomasWaldmann> yeah, but ugh X)
2007-09-23T01:10:27 <ThomasWaldmann> then maybe name has less effects
2007-09-23T01:10:29 <TheSheep> img has longdesc
2007-09-23T01:11:08 <ThomasWaldmann> that's maybe used by users
2007-09-23T01:11:45 <TheSheep> http://htmlhelp.com/reference/html40/attrs.html <-- these are common to all elements
2007-09-23T01:11:51 <ThomasWaldmann> i guess i just use title until we have something better
2007-09-23T01:13:30 <TheSheep> ThomasWaldmann: wait, this is html is only between fsckedit and html->wiki translator, right? it doesn't have to validate...
2007-09-23T01:13:47 <ThomasWaldmann> onlclick='foo="whatever"'
2007-09-23T01:14:28 <TheSheep> onclick="return True; foo='whatever'"
2007-09-23T01:14:45 <TheSheep> you don't want the links to stop working :)
2007-09-23T01:15:24 <ThomasWaldmann> the dont work in the editor anyway
2007-09-23T01:16:26 <ThomasWaldmann> well, I'll use title and leave that for later
2007-09-23T01:16:59 <ThomasWaldmann> cleaning up all that image/link mess and make roundtripping work is enough for now
2007-09-23T01:25:04 <ThomasWaldmann> TheSheep: how's 1.6 creole parser? :)
2007-09-23T01:25:34 <ThomasWaldmann> (maybe develop it in 1.7 branch first)
2007-09-23T01:25:43 <TheSheep> ThomasWaldmann: I have it opened on a second screen even as we speak
2007-09-23T01:25:55 <TheSheep> ThomasWaldmann: I think I even scrolled doen 3 or 4 lines this morning
2007-09-23T01:26:01 <TheSheep> down
2007-09-23T01:26:05 <ThomasWaldmann> |-)
2007-09-23T01:27:10 <TheSheep> I had to fight for disk space for it :D
2007-09-23T01:29:12 <ThomasWaldmann> disks are cheap those days
2007-09-23T01:29:25 <ThomasWaldmann> (sometimes too cheap)
2007-09-23T01:30:16 <TheSheep> you know how it is at the uni -- the chepest things are the most lacking, no problem buying the expensive ones
2007-09-23T01:30:36 <xorAxAx> ThomasWaldmann: not 2.5" disks :)
2007-09-23T01:30:45 <xorAxAx> 100 GB 96 EUR
2007-09-23T01:31:30 <TheSheep> maybe I should sell my 5ΒΌ" ones :)
2007-09-23T01:32:00 <TheSheep> but I doubt I'd even have 1GB
2007-09-23T01:32:11 <TheSheep> they are 360kB a piece
2007-09-23T01:35:52 <ThomasWaldmann> xorAxAx: even that is cheap
2007-09-23T01:36:17 <ThomasWaldmann> TheSheep: order some arrays and distribute the disk under the poor :D
2007-09-23T01:36:33 <TheSheep> ThomasWaldmann: there are no arrays fro floppies
2007-09-23T01:36:35 <TheSheep> for
2007-09-23T01:37:23 <ThomasWaldmann> floppies? what's that? :)
2007-09-23T01:40:16 <TheSheep> we still have the 8" floppies at the uni
2007-09-23T01:40:27 <xorAxAx> in poland they use RAIF setups
2007-09-23T01:40:27 <TheSheep> but no drive for them
2007-09-23T01:42:09 <TheSheep> xorAxAx: nah, for large amounts of data we have those drums with steel wire
2007-09-23T01:42:27 <xorAxAx> no toroidal core memory?
2007-09-23T01:42:46 <TheSheep> xorAxAx: you can record on them with this 500W electromagnet
2007-09-23T01:43:19 <TheSheep> xorAxAx: just like phonograph, only for data
2007-09-23T01:44:13 <xorAxAx> hmm
2007-09-23T01:44:40 <TheSheep> ibm uses something similar, only they use plastic tape
2007-09-23T01:44:43 <TheSheep> how fragile
2007-09-23T01:44:57 <TheSheep> a good, thick steel wire is so much better
2007-09-23T01:46:40 <TheSheep> if you spin the drums fast enough, you can power the whole computer just from the electromagnetic energy from the header
2007-09-23T02:01:59 <TheSheep> ok, I have text_creole.py from 1.7 running on 1.6
2007-09-23T02:02:11 <TheSheep> this was easy
2007-09-23T02:16:59 <ThomasWaldmann> TheSheep: we have no sane testing framework in 1.6 yet
2007-09-23T02:17:15 <ThomasWaldmann> so maybe develop it in 1.7 until it does everything you want
2007-09-23T06:56:26 <dreimark> gui needs to be fixed to use << >> for macros too
2007-09-23T08:25:20 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> default * 2855:f003d451be90 1.7/MoinMoin/converter/text_html_text_moin_wiki.py: text_html_text_moin_wiki: br for empty cell of OOo to gui replaced by whitespace at process_table_data
2007-09-23T08:25:23 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> default * 2856:604b5181385d 1.7/MoinMoin/converter/_tests/test_text_html_text_moin_wiki.py: test_text_html_text_moin_wiki: fixed testOOoTable01
2007-09-23T08:28:10 <CIA-27> moin: Reimar Bauer <rb.proj AT googlemail DOT com> default * 2201:a10d3240ff58 1.6/MoinMoin/converter/text_html_text_moin_wiki.py: text_html_text_moin_wiki: br for empty cell of OOo to gui replaced by whitespace at process_table_data
2007-09-23T08:28:47 <dreimark> ThomasWaldmann: I have replaced that ugly hack
2007-09-23T08:54:38 <dreimark> http://moinmo.in/MoinMoinBugs/1.6devGuiInsertingMacroUsesOldSyntax
2007-09-23T11:21:53 <TheSheep> ThomasWaldmann: good idea
2007-09-23T11:51:34 <TheSheep> um, how do you actually run the tests? :)
2007-09-23T11:52:46 <TheSheep> ok, found it :)
2007-09-23T12:18:13 <ThomasWaldmann> moin
2007-09-23T12:22:54 <TheSheep> ThomasWaldmann: moin
2007-09-23T12:23:06 <TheSheep> ThomasWaldmann: is http://moinmo.in/MoinDev/UnitTests up to date?
2007-09-23T12:23:34 <ThomasWaldmann> btw, a commit fixing some of the gui converter roundtripping problems will come later today
2007-09-23T12:24:17 <ThomasWaldmann> TheSheep: see the comment at top :)
2007-09-23T12:24:39 <TheSheep> ah, I somehow missed that, thanks
2007-09-23T12:25:20 <ThomasWaldmann> best is to look at some other test, that does not import unittest
2007-09-23T12:29:55 <TheSheep> ThomasWaldmann: yup, stealing test_text_moin_wiki.py now :)
2007-09-23T12:31:36 <ThomasWaldmann> TheSheep: before you commit (in general), try to run py.test from the toplevel dir
2007-09-23T12:51:55 <dreimark> bbl
2007-09-23T16:07:44 <TheSheep> argh, tests for pep8 compliance %)
2007-09-23T16:08:00 <TheSheep> you guys are evil ;)
2007-09-23T16:35:43 <ThomasWaldmann> :)
2007-09-23T16:36:26 <ThomasWaldmann> that saved me from lots of work for SOC2007 :)
2007-09-23T16:38:00 <ThomasWaldmann> what's still missing is a higher level "this code sucks" tester :D
2007-09-23T16:38:21 <ThomasWaldmann> but I guess we would get too much failures with that currently
2007-09-23T16:40:04 <dreimark> bb
2007-09-23T16:41:18 <ThomasWaldmann> hi dreimark
2007-09-23T16:41:54 <ThomasWaldmann> if you enter a<br>b in a table cell using the gui editor, what happens?
2007-09-23T16:44:56 <dreimark> ThomasWaldmann: ||<tablewidth="200px" tablestyle="">a<br>b || ||
2007-09-23T16:46:08 <dreimark> if I enter a<<BR>>b I fot a line break by BR
2007-09-23T16:47:22 * TheSheep trims all the lines up to 80 chars
2007-09-23T16:47:45 <TheSheep> having the editor window set to 80 characters wide helps a lot :)
2007-09-23T16:47:47 <ThomasWaldmann> TheSheep: ehrm?
2007-09-23T16:48:12 <ThomasWaldmann> please don't shorten lines to 80 chars (if they were just 90 or 100)
2007-09-23T16:48:36 <TheSheep> ThomasWaldmann: but I can't see them in my editor :(
2007-09-23T16:48:47 <TheSheep> 'sides, they were at least 150
2007-09-23T16:48:50 <ThomasWaldmann> that makes stuff often more hard to read
2007-09-23T16:49:20 <TheSheep> now I need to lear to understand what that py.test says to me :/
2007-09-23T16:49:30 <ThomasWaldmann> and we usually don't have terminals any more that are limited to 80
2007-09-23T16:50:03 <ThomasWaldmann> (ok for 150 chars stuff, of cours)
2007-09-23T16:50:17 <TheSheep> ThomasWaldmann: I have terminal windows set to 80c, because then I can have two of them side by side :)
2007-09-23T16:50:32 <dreimark> TheSheep: sometimes I do save the input and output to a file for comparison
2007-09-23T16:51:06 <ThomasWaldmann> TheSheep: if just DISLIKE code like this:
2007-09-23T16:51:41 <ThomasWaldmann> def foo(aaaaaa, bbbbb, ccccc, ...., eeee="foo",
2007-09-23T16:51:58 <ThomasWaldmann> ff="bar"):
2007-09-23T16:52:26 <TheSheep> :(
2007-09-23T16:52:40 <ThomasWaldmann> (just because 80chars was reached after "foo", )
2007-09-23T16:52:52 <TheSheep> well, luckily, no defs like that :D
2007-09-23T16:53:20 <TheSheep> but I have broken some regexps
2007-09-23T16:53:36 <ThomasWaldmann> (applies to all code, not just defs - I dislike using a new code line just for some few chars that could also go in the line above)
2007-09-23T16:53:41 <TheSheep> (in the sense that they are multiline now)
2007-09-23T16:53:48 <dreimark> TheSheep: on what are you on?
2007-09-23T16:53:55 <TheSheep> dreimark: the creole parser
2007-09-23T16:54:03 <ThomasWaldmann> re.X is fine for better understanding
2007-09-23T16:54:34 <TheSheep> I need to reformat those regexps first :/
2007-09-23T16:55:01 <ThomasWaldmann> maybe do it similar to the new wiki parser
2007-09-23T16:55:08 <TheSheep> I was thinking about putting all those rules to a separate class
2007-09-23T16:55:43 <TheSheep> ThomasWaldmann: yes, I already got rid of that _get_rule() thing and the inter_table
2007-09-23T16:56:19 <ThomasWaldmann> (don't copy the still broken/fscked stuff, though :)
2007-09-23T16:59:01 <dreimark> ThomasWaldmann: to be more clear the <br> gots escaped <tr><td>a<br>b</td><td></td></tr>
2007-09-23T16:59:27 <TheSheep> why are some docstrings broken at 55 characters?
2007-09-23T16:59:34 <ThomasWaldmann> dreimark: i am speaking of a br tag, not <br> just typed in
2007-09-23T16:59:50 <ThomasWaldmann> dreimark: maybe by pressing enter or shift-enter?
2007-09-23T17:00:11 <ThomasWaldmann> or copy&paste some table element from elsewhere
2007-09-23T17:01:06 <ThomasWaldmann> if you have some real table content that includes <br> tags, you don't just want to replace it by blanks, but generate <<BR>> macros for it
2007-09-23T17:02:02 <ThomasWaldmann> TheSheep: a nice thing we need sooner or later would be some tiny, separate link markup parser (maybe also transclusion)
2007-09-23T17:02:37 <ThomasWaldmann> because we use links at other places, too, e.g. quicklinks, navi_bar links
2007-09-23T17:03:00 <ThomasWaldmann> iirc, they currently all have their own, small, broken code to do that
2007-09-23T17:03:01 <TheSheep> ThomasWaldmann: the best thing would be to somehow have the link markup parser separate, with the same code somehow included and used in the "large" parsers
2007-09-23T17:03:45 <TheSheep> note that you can't really parse "only links", you need to llok at things like {{{ }}} too
2007-09-23T17:04:06 <ThomasWaldmann> (at least those sharing the link markup, probably only wiki and creole and those other uses)
2007-09-23T17:04:38 <ThomasWaldmann> but I guess we should delay that until after 1.6 release
2007-09-23T17:04:46 <TheSheep> definitely :)
2007-09-23T17:07:49 <dreimark> ThomasWaldmann: this is a different problem as the one I've fixed
2007-09-23T17:08:14 <dreimark> I do investigate that one too
2007-09-23T17:09:24 <ThomasWaldmann> why is it different?
2007-09-23T17:11:02 <dreimark> the table line gots broken because of a newline \n.
2007-09-23T17:11:49 <ThomasWaldmann> well, that's all related
2007-09-23T17:12:30 <ThomasWaldmann> of course you don't want to break table cells by using \n, but you also don't want to throw away all <br> by just replacing them with blank
2007-09-23T17:13:10 <ThomasWaldmann> so just replace them with <<BR>> and after that, maybe optimize the case when there is ONLY <<BR>> in a table cell and replace that by a blank
2007-09-23T17:14:21 <dreimark> I do agree that it is saver to use <<BR>> instead of whitespace.
2007-09-23T17:14:42 <dreimark> But this does not change anything for the test you asked for.
2007-09-23T17:15:25 <ThomasWaldmann> sure it does
2007-09-23T17:15:46 <ThomasWaldmann> your current code just replaces all <br> tags by blanks
2007-09-23T17:16:13 <dreimark> no
2007-09-23T17:16:42 <dreimark> but I try again
2007-09-23T17:17:13 <ThomasWaldmann> + if name == 'br':
2007-09-23T17:17:13 <ThomasWaldmann> + # if we get a br for an empty cell from e.g. cut and paste from OOo
2007-09-23T17:17:13 <ThomasWaldmann> + # to the gui it should be appended as white_space.
2007-09-23T17:17:13 <ThomasWaldmann> + self.text.append(self.white_space)
2007-09-23T17:17:13 <ThomasWaldmann> + found = True
2007-09-23T17:17:19 <ThomasWaldmann> no?
2007-09-23T17:17:41 <ThomasWaldmann> and the comment is wrong
2007-09-23T17:18:00 <ThomasWaldmann> it is what you wanted to do, buy not what it does, IMHO
2007-09-23T17:19:28 <dreimark> I will make it clearer
2007-09-23T17:19:43 <dreimark> but the case you wanted solved is not affected by that line
2007-09-23T17:19:53 <ThomasWaldmann> why?
2007-09-23T17:19:57 <dreimark> this is broken in 1.5.8 too
2007-09-23T17:20:32 <dreimark> it makes a real \n ||<tablewidth="200px" tablealign="">a \nb|| ||
2007-09-23T17:20:47 <dreimark> where it should write a <<BR>>
2007-09-23T17:21:08 <ThomasWaldmann> ?
2007-09-23T17:21:18 <dreimark> http://moinmo.in/ReimarBauer/Test2
2007-09-23T17:21:21 <ThomasWaldmann> lets talk on the html -> wiki markup level
2007-09-23T17:21:46 <ThomasWaldmann> <td>a<br>b</td> -> ||a<<BR>>b||
2007-09-23T17:21:52 <ThomasWaldmann> that should work
2007-09-23T17:22:52 <ThomasWaldmann> <td><br></td> -> || || this is a post-processed special case (due to browser brokenness)
2007-09-23T17:23:22 <TheSheep> ThomasWaldmann: why [^\s] in the escape_rule ?
2007-09-23T17:23:43 <TheSheep> ThomasWaldmann: (sorry, nevermind)
2007-09-23T17:25:18 <dreimark> gui sourcecode for the first example <tr><td>a<br />b</td></tr>
2007-09-23T17:25:51 <dreimark> in wiki markup ||<tablewidth="200px" tablealign="">a\nb||
2007-09-23T17:26:14 <ThomasWaldmann> yeah, that's obviously wrong
2007-09-23T17:26:20 <dreimark> there is a linebreak before b (no whitesapce and if changed no <BR>
2007-09-23T17:27:02 <dreimark> so it goes to somewhere else but not through that part of code.
2007-09-23T17:28:10 <ThomasWaldmann> then the fix is maybe still at the wrong place
2007-09-23T17:28:51 <ThomasWaldmann> btw, I suggest you make some tests testing specifically this behaviour and commit when they pass
2007-09-23T17:29:26 <ThomasWaldmann> (so we won't have all those commits/reverts/fixes... in the repo)
2007-09-23T17:30:51 <dreimark> yeah, sorry for this mess, (but I have witten tests always)
2007-09-23T17:31:34 <ThomasWaldmann> and try to do some more minimal tests. while these OOo tests are maybe nice for real-world testing, some more simple tests would better show the problems.
2007-09-23T17:32:30 <dreimark> I got that too, fixing those real world tests is a mess too
2007-09-23T17:33:14 <dreimark> Do you look at this eating of text in front of two macros?
2007-09-23T17:33:52 <ThomasWaldmann> no, I am currently cleaning up the link/transclusion mess in gedit formatter / converter
2007-09-23T17:35:02 <ThomasWaldmann> I am almost finished, but some magic still needs treatment.
2007-09-23T17:35:58 <ThomasWaldmann> [[target|{{image}}]] I made some magic that puts target into the title/alt of image
2007-09-23T17:36:26 <ThomasWaldmann> so when you move the mouse over the image, you see where you are going
2007-09-23T17:37:05 <ThomasWaldmann> but currently, the gui converter translates this back to [[target|{{image|target}}]] which is unwanted
2007-09-23T17:38:06 <dreimark> arrgh
2007-09-23T17:39:26 <ThomasWaldmann> btw, GetVal tests still failing.
2007-09-23T17:39:58 <ThomasWaldmann> brb
2007-09-23T17:42:59 <dreimark> not here - strange
2007-09-23T18:12:42 <TheSheep> hmm, moin-1.7 seems to still use moin_static160 prefix
2007-09-23T18:23:56 <dreimark> Yippee!
2007-09-23T18:25:14 <ThomasWaldmann> :)
2007-09-23T18:26:21 <dreimark> its the right place, one line more, now new tests doing
2007-09-23T18:41:10 <TheSheep> ok, looks like I've broken it hard enough that it works :)
2007-09-23T19:09:51 <TheSheep> the pep8 test complain about trailing whitespace inside *strings literals* :(
2007-09-23T19:39:28 <ThomasWaldmann> yes it does :)
2007-09-23T19:39:44 <ThomasWaldmann> it is not too intelligent, but otoh, where do you need that?
2007-09-23T19:46:13 * TheSheep runs :%s/\s\+$//g
2007-09-23T19:46:16 <TheSheep> grumble
2007-09-23T19:59:34 <ThomasWaldmann> the src test has a switch to do that
2007-09-23T20:03:41 <TheSheep> ThomasWaldmann: what is the syntax for macros in the moin parser? <<MacroName|args>> or <<MacroName(args)|alt text>> ?
2007-09-23T20:03:48 <TheSheep> ThomasWaldmann: or something else?
2007-09-23T20:04:58 <TheSheep> hmm... looks like <<MacroName args>>
2007-09-23T20:05:37 <TheSheep> ah, no
2007-09-23T20:05:44 <TheSheep> <<MacroName(args)>>
2007-09-23T20:06:01 <TheSheep> so I'll make it <<MacroName(args)|alt>> in creole :)
2007-09-23T20:08:36 <dreimark> ThomasWaldmann: did you have tried smileys in gui ?
2007-09-23T20:09:43 <ThomasWaldmann> dreimark: not yet
2007-09-23T20:09:56 <ThomasWaldmann> TheSheep: what should that be?
2007-09-23T20:10:19 <ThomasWaldmann> (the macro code can emit an alt attr if it wants, derived from a param)
2007-09-23T20:10:43 <dreimark> http://test17.wikiwikiweb.de/ReimarBauer/smiley Very strange result
2007-09-23T20:10:52 <TheSheep> ThomasWaldmann: well, but will it be descriptive enough? :)
2007-09-23T20:12:30 <ThomasWaldmann> it can be as descriptive (and easier) as |alt
2007-09-23T20:12:52 <ThomasWaldmann> better hack the links and transclusion stuff with |params :)
2007-09-23T20:13:24 <dreimark> ahh the url is not complete {{/wiki/modern/img/angry.png}}
2007-09-23T20:13:43 <ThomasWaldmann> btw, I put johill's arg parser in 1.6 wikiutil.py, so one can use it for params
2007-09-23T20:13:55 <ThomasWaldmann> dreimark: that IS complete
2007-09-23T20:14:57 <ThomasWaldmann> dreimark: and btw, I am currently working on that image and link stuff, so please dont make conflicts
2007-09-23T20:15:05 <dreimark> then is my browser broken -> Der Server unter wiki konnte nicht gefunden werden.
2007-09-23T20:15:48 <dreimark> I dont work on that I have to solve strong in tables.
2007-09-23T20:30:42 <TheSheep> I don't get the testEscapeInGetTextFormatted test :/
2007-09-23T20:30:59 <TheSheep> I'm getting double escaping
2007-09-23T20:33:49 <TheSheep> am I supposed to somehow detect if the input is already escaped?
2007-09-23T20:38:25 <dreimark> weird, OOo formats its blank sign of BR too. LOL a strong BR or italic
2007-09-23T20:55:19 <dreimark> ThomasWaldmann: that adding of the arg parser is great
2007-09-23T20:56:31 <TheSheep> now we only need to rewrite all the macros to use it :)
2007-09-23T20:59:49 <dreimark> TheSheep: its not necessary to do it now the old style works too
2007-09-23T21:01:36 <ThomasWaldmann> TheSheep: that rewrite happens in 1.7+
2007-09-23T21:03:17 <TheSheep> ok, another test that I completely don't understand
2007-09-23T21:03:30 <TheSheep> testEmptyLineBeforeBlock
2007-09-23T21:04:01 <TheSheep> first it states that there should not be an editional <p> in there, then it makes sure there is one O.io
2007-09-23T21:04:05 <TheSheep> O.o
2007-09-23T21:04:28 <TheSheep> additional
2007-09-23T21:10:47 <TheSheep> stupid question
2007-09-23T21:11:01 <TheSheep> why is formatter.code() replacing all the spaces with ?
2007-09-23T21:11:17 <TheSheep> you could get the same result much easier with CSS
2007-09-23T21:14:42 * TheSheep replaces it with rawHTML
2007-09-23T21:16:31 <dreimark> http://test17.wikiwikiweb.de/ReimarBauer/TableTests
2007-09-23T21:16:37 <dreimark> did I missed one
2007-09-23T22:16:22 <TheSheep> yay! =========== tests finished: 823 passed, 66 skipped in 61.71 seconds ============
2007-09-23T22:19:57 <TheSheep> now, what do I do with the bundle?
2007-09-23T22:23:30 <ThomasWaldmann> hg export <changeset> and send me your ssh pubkey :)
2007-09-23T22:35:29 <TheSheep> sent :)
2007-09-23T22:35:43 <TheSheep> the paths are huge, I will commit more often now
2007-09-23T22:35:54 <TheSheep> patches
2007-09-23T22:39:23 * ThomasWaldmann looks into mail
2007-09-23T22:40:03 <ThomasWaldmann> dreimark: the GetVal stuff succeeds for first py.test run and fails for 2nd (run from toplevel moin dir)
2007-09-23T22:40:16 <ThomasWaldmann> (first after nuking the test wiki)
2007-09-23T22:40:35 <TheSheep> ThomasWaldmann: it appears in every other test for me
2007-09-23T22:41:30 <ThomasWaldmann> heh
2007-09-23T22:41:53 <dreimark> nuked, and py.test started
2007-09-23T22:45:25 <TheSheep> is there a way to get a real </p> with the html formatter?
2007-09-23T22:45:36 <TheSheep> other than rawHTML
2007-09-23T22:46:08 <dreimark> first test of test_packages.py fails
2007-09-23T22:47:02 <ThomasWaldmann> TheSheep: try a fresh clone of 1.7, apply your patches and push to 1.7 main repo by ssh
2007-09-23T22:47:44 <TheSheep> ThomasWaldmann: ok
2007-09-23T22:48:15 <ThomasWaldmann> please use GOOD comments
2007-09-23T22:48:36 <ThomasWaldmann> e.g. start with "Creole parser: ..."
2007-09-23T22:49:28 <dreimark> ThomasWaldmann: did testBasicPackageThing fail for you too?
2007-09-23T22:49:51 <TheSheep> ThomasWaldmann: ok, I'm sorry, I didn't look at other comments before commiting, I should have
2007-09-23T22:50:10 <ThomasWaldmann> and please add that rsplit like it was before
2007-09-23T22:50:17 <dreimark> ThomasWaldmann: test_GetVal doesnt fail for me
2007-09-23T22:50:50 <ThomasWaldmann> TheSheep: or it will fail on 2.3 (we still support py 2.3)
2007-09-23T22:53:06 <ThomasWaldmann> TheSheep: + if word.startswith('..%s' % wikiutil.CHILD_PREFIX):
2007-09-23T22:53:16 <ThomasWaldmann> how about PARENT_PREFIX?
2007-09-23T22:55:19 <TheSheep> ThomasWaldmann: right away!
2007-09-23T22:57:18 <dreimark> MoinMoin/_tests/test_packages.py fails always since I have purged the test wiki
2007-09-23T22:59:12 * ThomasWaldmann tries
2007-09-23T22:59:55 <ThomasWaldmann> cd moin-1.7
2007-09-23T22:59:58 <ThomasWaldmann> py.test
2007-09-23T23:00:14 <TheSheep> ok, I'm pushing
2007-09-23T23:00:17 <ThomasWaldmann> -> = tests finished: 383 passed, 1 failed, 69 skipped in 23.04 seconds =
2007-09-23T23:00:29 <CIA-27> moin: Radomir Dopieralski <moindev@sheep.art.pl> default * 2857:b5b3a9d16abf 1.7/MoinMoin/parser/text_creole.py: Creole parser: removed _get_rule() and inter_table
2007-09-23T23:00:32 <CIA-27> moin: Radomir Dopieralski <moindev@sheep.art.pl> default * 2858:70a1f88575ce 1.7/MoinMoin/parser/ (text_creole.py _tests/test_text_creole.py): Creole parser: tests for creole parser
2007-09-23T23:00:33 <CIA-27> moin: Radomir Dopieralski <moindev@sheep.art.pl> default * 2859:2087924a60f1 1.7/MoinMoin/parser/text_creole.py: Creole parser: put back rsplit needed for python 2.3
2007-09-23T23:00:34 <CIA-27> moin: Radomir Dopieralski <moindev@sheep.art.pl> default * 2860:60f95d2b090b 1.7/MoinMoin/parser/text_creole.py: Creole parser: replaced hardcoded ../ with PARENT_PREFIX in interwiki
2007-09-23T23:01:17 <TheSheep> it works! \o/
2007-09-23T23:01:22 <ThomasWaldmann> TheSheep: welcome to core team group :)
2007-09-23T23:01:39 <TheSheep> ThomasWaldmann: thank you
2007-09-23T23:01:44 * TheSheep cries a little
2007-09-23T23:02:30 * ThomasWaldmann gives TheSheep a hanky
2007-09-23T23:05:52 <dreimark> TheSheep: welcome too
2007-09-23T23:08:42 <dreimark> ThomasWaldmann: for me is always this test failing testBasicPackageThings
2007-09-23T23:08:44 <dreimark> tests finished: 391 passed, 1 failed, 69 skipped in 195.96 seconds
2007-09-23T23:09:22 <ThomasWaldmann> we need to remove the random() from the tests :)
2007-09-23T23:09:45 <ThomasWaldmann> for me it is alternating with failing/succeeding the getval stuff
2007-09-23T23:10:37 <dreimark> this happens if somewhere the gobal state is changed before that test
2007-09-23T23:11:44 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2861:e1ce3040fcd6 1.7/MoinMoin/formatter/text_gedit.py: gedit formatter: remove special casing for ImageLink
2007-09-23T23:11:46 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2862:8c1d40768df7 1.7/MoinMoin/ (3 files in 3 dirs): gedit formatter/converter: fix links/transclusion (suppressing default alt texts needs more work)
2007-09-23T23:11:47 <CIA-27> moin: Thomas Waldmann <tw AT waldmann-edv DOT de> default * 2863:244f9e89fb15 1.7/MoinMoin/ (4 files in 4 dirs): merged main
2007-09-23T23:15:06 <dreimark> TheSheep: you should obfuscate your email
2007-09-23T23:15:33 <TheSheep> dreimark: nah
2007-09-23T23:16:14 <TheSheep> dreimark: I don't believe in spam :3
2007-09-23T23:17:34 <dreimark> the filters got better I remember something about 200 spam mails / day
2007-09-23T23:17:45 <ThomasWaldmann> at least the hgweb does not make mailto: links out of it
2007-09-23T23:17:51 <TheSheep> dreimark: graylisting helps a lot
2007-09-23T23:18:07 <dreimark> and blacklist too
2007-09-23T23:18:58 <dreimark> ThomasWaldmann: more ideas http://test17.wikiwikiweb.de/ReimarBauer/TableTests
2007-09-23T23:20:07 <ThomasWaldmann> yeah, have one with <br>
2007-09-23T23:21:32 <ThomasWaldmann> ||a||<<BR>>|| while this is ok for the beginning, we want to optimize away that <<BR>>
2007-09-23T23:21:44 <dreimark> testTable03
2007-09-23T23:21:47 <ThomasWaldmann> (and replace it by a blank if there is nothing else in that cell)
2007-09-23T23:22:01 <dreimark> ||a<<BR>>b||
2007-09-23T23:22:21 <dreimark> ThomasWaldmann: that optimisation could be done in a further step
2007-09-23T23:23:19 <ThomasWaldmann> ''' is strong? not b?
2007-09-23T23:24:29 <dreimark> '''is strong
2007-09-23T23:25:58 <ThomasWaldmann> right, we generate strong (and use is_b to track it :-P)
2007-09-23T23:26:52 <TheSheep> there was a nice rant somewhere on the wab about using <strong> when you really mean <b>
2007-09-23T23:27:25 <dreimark> OOo wants to have '''<<BR>>''' if the cell is formatted bold and empty
2007-09-23T23:27:41 <ThomasWaldmann> pfff
2007-09-23T23:28:04 <ThomasWaldmann> how does empty boldness look?
2007-09-23T23:28:12 <dreimark> empty
2007-09-23T23:28:21 * ThomasWaldmann guessed so
2007-09-23T23:28:35 <dreimark> and if its saved br is replaced by \n
2007-09-23T23:29:02 <dreimark> because of recursive call in process_inline
2007-09-23T23:29:56 <dreimark> fckeditor does not add formatcodes to empty cells
2007-09-23T23:32:22 <dreimark> so it does tangent only insertion of OOo (formatted blanks)
2007-09-23T23:33:04 <ThomasWaldmann> try word and excel to see more crap :P
2007-09-23T23:34:27 <dreimark> planned for the next days
2007-09-23T23:34:54 <dreimark> hmm, http://localhost:8080/FrontPage?mimetype=text/plain
2007-09-23T23:35:07 <dreimark> seems to be enough to call a formatter
2007-09-23T23:35:23 <dreimark> is the parameter formatter obsolete now?
2007-09-23T23:40:59 <ThomasWaldmann> dreimark: try it. that stuff will need more design when we have mimetype items
2007-09-23T23:41:45 <dreimark> it works that way, but I am not sure if its a feature or bug.
2007-09-23T23:42:22 <dreimark> was it possible in the past to use another mimetype as the formatter call
2007-09-23T23:42:51 <dreimark> http://localhost:8080/FrontPage?formatter=text_plain&mimetype=text/csv
2007-09-23T23:43:15 <dreimark> fails now because I have no text_csv formatter
2007-09-23T23:43:48 <ThomasWaldmann> what should be the sense of that?
2007-09-23T23:56:16 <dreimark> if you do that with your browser you have a definition for a mimetype for an application
2007-09-23T23:56:55 <dreimark> to handle the output directly
2007-09-23T23:57:29 <dreimark> text/csv opens OOo for example
2007-09-23T23:57:42 <ThomasWaldmann> so you need a mimetype override for the text/plain formatter
2007-09-23T23:58:22 <ThomasWaldmann> did you try #!format csv ?
2007-09-23T23:59:04 <dreimark> no, it was an example I tried to write a small example for a formatter and came to that
MoinMoin: MoinMoinChat/Logs/moin-dev/2007-09-23 (last edited 2007-10-29 19:16:49 by localhost)