MoinMoin: MoinMoinChat/Logs/moin-dev/2007-01-06

2007-01-06T10:52:52  <nomed> hi all
2007-01-06T10:53:16  <xorAxAx> hi
2007-01-06T10:53:44  <nomed> first of all a question ..
2007-01-06T10:54:00  <nomed> xorAxAx: do u have the sessinParser.py in front of you ?
2007-01-06T10:54:35  <xorAxAx> yes
2007-01-06T10:54:40  <nomed> ok
2007-01-06T10:54:53  <nomed> parseSession returns a dict
2007-01-06T10:55:42  <nomed> i couldn't see the one returned parsing an egw session
2007-01-06T10:56:10  <nomed> but what it was unclear was:
2007-01-06T10:56:21  <nomed> def handle_egroupware(session):
2007-01-06T10:56:32  <nomed> in auth.php_session
2007-01-06T10:56:39  <xorAxAx> yes?
2007-01-06T10:56:53  <nomed> username = session['egw_session']['session_lid'].split("@", 1)[0]
2007-01-06T10:56:58  <nomed> in particular
2007-01-06T10:57:30  <nomed> i'd expect:
2007-01-06T10:57:50  <nomed> username = session[key]
2007-01-06T10:57:54  <xorAxAx> ?
2007-01-06T10:58:13  <xorAxAx> egw_session is the session name in php
2007-01-06T10:58:17  <xorAxAx> session_lid is the var name
2007-01-06T10:58:24  <xorAxAx> and then its parsed ...
2007-01-06T10:58:32  <xorAxAx> by split("@", 1)[0]
2007-01-06T10:59:13  <nomed> let's suppose a string like : login_ok|b:1;nome|s:4:"sica";inteiro|i:34;
2007-01-06T10:59:48  <nomed> what's the dict you should get ?
2007-01-06T11:00:00  <nomed> i ask this because at the moment ..
2007-01-06T11:00:11  <nomed> when i parse the object session
2007-01-06T11:00:53  <nomed> i check if it's an "O"
2007-01-06T11:01:09  <nomed> and then i consider the string between {}
2007-01-06T11:02:03  <nomed> http://phpfi.com/190943
2007-01-06T11:02:03  <xorAxAx> can i see your session blob again?
2007-01-06T11:03:31  <nomed> http://phpfi.com/190944
2007-01-06T11:04:48  <nomed> i could surely have something like:
2007-01-06T11:05:16  <xorAxAx> if you do it correctly, then your little function will access session like this:
2007-01-06T11:05:17  <nomed> session_dict = {'UlteoSession':{key:value}}
2007-01-06T11:05:31  <xorAxAx> session["ULTEO_SESS"]["UlteoSession"]
2007-01-06T11:05:43  <xorAxAx> and from that dict, it will fetch the items
2007-01-06T11:05:52  <xorAxAx> login and displayname
2007-01-06T11:06:58  <nomed> ok i'm going to fix it that way
2007-01-06T11:07:17  <xorAxAx> the outer dict is generated by parseSession
2007-01-06T11:07:28  <xorAxAx> and the inner stuff is done by parseValue
2007-01-06T11:07:28  <nomed> at the moment i fixed it in a way i could get it working for my needs
2007-01-06T11:08:17  <nomed> ok
2007-01-06T11:11:23  <nomed> other question is:
2007-01-06T11:11:39  <nomed> even if the user has a session enabled
2007-01-06T11:11:57  <nomed> the user should create the account in moin
2007-01-06T11:12:10  <nomed> correct ?
2007-01-06T11:13:15  <nomed> user.create_or_update(changed) in auth.php_session?
2007-01-06T11:14:09  <nomed> that should just update an account in case it has been changed
2007-01-06T11:14:17  <nomed> or it seems that's waht happens
2007-01-06T11:15:10  <nomed> did i understand correctly ?
2007-01-06T11:16:05  <xorAxAx> yes, it will update it if it changed
2007-01-06T11:17:10  <nomed> shouldn't it create the user in case it doesn't exist
2007-01-06T11:17:19  <nomed> ?
2007-01-06T11:23:16  <xorAxAx> yes, it does
2007-01-06T11:27:34  <nomed> xorAxAx: any idea what it can be if it doesn't ?
2007-01-06T11:28:24  <xorAxAx> you have to set a config setting, nomed
2007-01-06T11:28:30  <xorAxAx> see HelpOnConfiguration
2007-01-06T11:28:56  <nomed> ok
2007-01-06T12:26:28  <xorAxAx> gaaa, referencing changesets by ordinal numbers is pretty pointless ...
2007-01-06T12:26:38  <xorAxAx> please use short hashes
2007-01-06T12:28:13  <dreimark> ?
2007-01-06T12:29:01  <xorAxAx> dreimark: every changeset in hg has 2 ids - a revision number and a hash
2007-01-06T12:29:15  <xorAxAx> dreimark: the revision id is different on every computer, though
2007-01-06T12:29:35  <xorAxAx> so if somebody writes "reverting changeset 123" it is rather pointless
2007-01-06T12:29:54  <xorAxAx> in order to avoid this confusion, use the short hash
2007-01-06T12:30:37  <ThomasWaldmann> dreimark: i have a look at the xapian problem
2007-01-06T12:37:14  <nomed> http://www.ulteo.com/~nomed/sessionParser.py
2007-01-06T12:37:22  <nomed> xorAxAx: take a look please
2007-01-06T12:37:35  <nomed> moin 1.5
2007-01-06T12:38:18  <xorAxAx> nomed: i wont accept it for 1.5
2007-01-06T12:38:42  <nomed> xorAxAx: that's ok for me
2007-01-06T12:38:52  <nomed> but i wrote it for the wiki i'm using
2007-01-06T12:39:04  <nomed> i guess it should be the same for 1.6
2007-01-06T12:39:16  <xorAxAx> yeah, sure
2007-01-06T12:39:33  <xorAxAx> hmm, it doesnt work if some of your object data stuff has "{" in it
2007-01-06T12:39:46  <xorAxAx> --> dont search for the braces :)
2007-01-06T12:40:12  <nomed> infact it could be a +3
2007-01-06T12:41:21  <ThomasWaldmann> dreimark: request.html_formatter should be there and initialized. we have to find out why it is not. but i have to go now, maybe I have time to look at it this weekend.
2007-01-06T12:42:28  <ThomasWaldmann> bbl
2007-01-06T16:19:57  <dreimark> ThomasWaldmann:that xapian bug happens with a fresh new instance too
2007-01-06T16:20:43  <dreimark> I was not able to install  PyStemmer - compile failure - but it should go without that
2007-01-06T16:34:05  <ThomasWaldmann> py 2.5?
2007-01-06T16:36:58  <dreimark> Python 2.4.4 (#2, Oct 20 2006, 00:23:25)
2007-01-06T16:50:21  <ThomasWaldmann> hmm, sollte eigentlich tun