[07:51] <dholbach> good morning
=== davidcalle_afk is now known as davidcalle
[08:41] <davidcalle> mhall119: wow, very nice :)
[09:02] <davidcalle> dholbach: do you want a second pair of eyes on the links bug?
[09:02] <dholbach> davidcalle, I'll put together a minimal test case and let you know
[09:02] <dholbach> thanks!
[09:03] <davidcalle> ok :à)
[09:49] <dholbach> davidcalle, http://pad.ubuntu.com/oRcuTY5ZJ3
[09:49] <dholbach> sorry it took a bit longer to put it together
[09:49] <dholbach> the test case is actually quite close to what I test on the live system, so I'm not quite sure yet what the issue is :)
[10:07]  * davidcalle looks
[10:14] <dholbach> davidcalle, when I add some print() statements in article.rewrite_links(), it looks like the link is changed, but maybe it's not saved correctly?
[10:14] <davidcalle> dholbach: still branching pips... :p
[10:14] <dholbach> sure sure, take your time
[10:58] <davidcalle> dholbach: that's weird
[10:58] <dholbach> I think I'm getting closer
[10:59] <davidcalle> dholbach: looks like the self.html changed in replace_links is not the one used at the publication step
[11:00] <dholbach> yep, I'm looking into that now
[11:00] <dholbach>  /o\
[11:19] <dholbach> davidcalle, ok, looks like that was it - I'll need to fix a few bits around it again, but I'm on the right track
[11:19] <dholbach> thanks for your help
[11:19] <dholbach> we arrived at the same conclusion, at the same time :-)
[11:52] <dholbach> lunch time, bbiab
[12:05] <redwolf> morning all
[12:05] <redwolf> anybody having problems with wiki.ubuntu.com?
[13:10] <dholbach> davidcalle, I pushed a bunch of changes the the links are now correct... a new problem I'm seeing now is a broken test and some escaped html, ie &lt;body&gt;
[13:10] <dholbach> but I'll figure that out
[13:10] <davidcalle> dholbach: escaped html would suggest the doc is passed twice through the md->html function
[13:13] <dholbach> I'll take a look, it could also be something in beautifulsoup(?) or the clean_html from djangocms_text_ckeditor.html we use
[13:13] <dholbach> just wanted to let you know how things are going
[13:18] <dholbach> ah, I know what it is - we prettify the HTML, so our regex in _remove_body_and_html_tags is broken :)
[13:18] <davidcalle> Oh right!
[13:19] <dholbach> getting there :-D
[13:25] <davidcalle> dholbach: one thing I've noticed after pulling your branch, replaced links should start with a "/"
[13:27] <dholbach> ok... I'll note this down
[14:12] <dholbach> it was beautifulsoup's prettify :)
[14:12] <dholbach> now the broken test
[16:59] <dholbach> davidcalle, the last issue I'm chasing is that a second import (with same content) seems to result in a new publish cycle (which would break the translations dashboard as we discussed before)
[16:59] <dholbach> I'm closer already, but haven't 100% figure it out yet :-/
[17:01] <davidcalle> dholbach: bzr (well, git) status and only work on modified and new docs? (just throwing out the first idea that comes)