[Pywikipedia-l] Strange month names from Site.mediawiki_message

[Bináris]

Good morning!

I collected month names at
http://hu.wikipedia.org/wiki/Szerkeszt%C5%91:Bin%C3%A1ris/Tartalomjegyz%C3%A9kbot/H%C3%B3napok_%E2%80%93_Months
as shown in archivebot.py. I need it for writing and testing the
TOCbot<http://hu.wikipedia.org/wiki/Szerkeszt%C5%91:Bin%C3%A1ris/TOCbot>,
which has to recognize short and long month names in archive name patterns.
The relevant part of my code is:
*    text += '<br>'.join([Site.mediawiki_message(MonthsS[i]) for i in
range(12)]) + '\n|'
    text += '<br>'.join([Site.mediawiki_message(MonthsL[i]) for i in
range(12)])
*for each Site in site.validLanguageLinks().

If there is any Dutch here, please check the *nl* row, because I saw that
archivebot.py handles it in a specific way that I don't understand, and
cannot check because of different locale.

Now, some rows show irregular behavior which I don't understand.

   1. In rows *ne* and *ti* a part of the short names seems to be native,
   and some of the short names English. (Apr, Oct in *ne* and from Jul to
   Dec in *ti*.) Is that perhaps a MediaWiki bug?
   2. In the last row, *zu*, all the long names have a 'u' at the beginning,
   just as in Python Unicode strings. Is that correct or some misbehaviour?
   3. In the row *ba* there are two short names (ғин, май) that don't fit in
   the line. Or is that the other ten? They are different and lower case, while
   the others upper case and identical to the long name. Is that a bug or a
   feature? If bug, is it in MediaWiki?
   4. There are two rows, *ba* and *ky*, where there are alternative
   versions in parenthesis, the basic version being *almost* identical to
   Russian. Does it mean that any of the given versions may appear in that wiki
   as a month name?

-- 
Bináris
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.wikimedia.org/pipermail/pywikipedia-l/attachments/20110425/96139a18/attachment.htm