1 2013-09-06T07:58:18 <ThomasWaldmann> moin
2 2013-09-06T10:54:09 <ananasova> moin
3 2013-09-06T11:07:54 <ananasova> ThomasWaldmann: is diff(old_text, new_text) better?
4 2013-09-06T11:09:30 <ThomasWaldmann> ananasova: can you just look yourself how the other api calls are named and just choose same convention except there is anything speaking against that?
5 2013-09-06T11:09:57 <ThomasWaldmann> (and same for params)
6 2013-09-06T11:16:44 <ananasova> most of the api methods that deal with diffs get revs as params. but we need something more general, because after destroy there are no revisions
7 2013-09-06T11:21:35 <ThomasWaldmann> ok, then think about something more general.
8 2013-09-06T11:21:52 <ThomasWaldmann> if it is called "old_text" it doesn't look general
9 2013-09-06T11:26:19 <ThomasWaldmann> (the reason why i was using revs there is that data might be large and revs have file-like api for data)
10 2013-09-06T11:27:36 <ananasova> then maybe use a string buffer StringIO?
11 2013-09-06T11:28:05 <ThomasWaldmann> you want to read that iso image into memory?
12 2013-09-06T11:36:34 <ananasova> anyway we need to read the whole contents into memory to construct the diff. or use generators
13 2013-09-06T11:41:22 <ThomasWaldmann> well, for some stuff, like iso images, you won't compute a diff
14 2013-09-06T11:41:58 <ThomasWaldmann> in fact, for most non-text stuff
15 2013-09-06T11:46:53 <ananasova> well, while building the notification message, no matter what is the item type, we first read the data and call the method _get_data_diff_text(), but only the text-like data will be used to compute the diff
16 2013-09-06T11:47:25 <ananasova> maybe we shouldn't read it at all
17 2013-09-06T11:47:41 <ThomasWaldmann> sure
18 2013-09-06T11:48:48 <ThomasWaldmann> reading should be done (if it makes sense) by the method (that can be implemented differently for each type)
19 2013-09-06T11:49:53 <ananasova> ok, now i see where i was blind. fixing it
20 2013-09-06T12:45:35 *** dave_largo
21 2013-09-06T13:02:09 <ananasova> ThomasWaldmann: please review when free https://codereview.appspot.com/13418045/
22 2013-09-06T15:10:55 <ThomasWaldmann> ananasova: done
23 2013-09-06T17:28:06 <ananasova> ThomasWaldmann: updated https://codereview.appspot.com/13418045/ , not sure about the docstring from the base Content class
24 2013-09-06T18:13:26 <ananasova> also updated cr 'Create notification message' - https://codereview.appspot.com/12745048/
25 2013-09-06T19:52:31 <ananasova> is there an easy way (without running a query) to find the name of the item based on it's itemid?
26 2013-09-06T20:59:04 *** dave_largo
27 2013-09-06T21:34:39 <ThomasWaldmann> note: pyquery for validation instead of lxml?
28 2013-09-06T21:35:27 <ThomasWaldmann> ananasova: for last question: no (except if you already have all metadata or name from url)
29 2013-09-06T21:54:45 <ananasova> the plan was to move to elementtree and not use lxml, no?
30 2013-09-06T21:58:29 <ananasova> i mean how is pyquery better than elementtree?
31 2013-09-06T22:05:16 <ananasova> gn
32 2013-09-06T23:05:48 *** glenn
33 2013-09-06T23:06:23 <glenn> I've installed Moin2 under Linux several times now with a high level of success. Wondering if I could provide anything useful with regard to my methods and results
34 2013-09-06T23:06:47 *** glenn
35 2013-09-06T23:07:19 *** glenn
36 2013-09-06T23:14:55 *** glenn
37
MoinMoin: MoinMoinChat/Logs/moin-dev/2013-09-06 (last edited 2013-09-06 08:00:03 by IrcLogImporter)