On Wed, Mar 11, 2009 at 02:50:11PM +0100, Michal Čihař wrote: > > What is correct according to upstream, your application should take > care of conversion unicode sting to whatever is proper representation > of it on filesystem. >
Oh. Strange. That illusion seems to be supported by the Python itself: $ python Python 2.5.2 (r252:60911, Jan 4 2009, 21:59:32) [GCC 4.3.2] on linux2 Type "help", "copyright", "credits" or "license" for more information. >>> f = open(u"łä.mp3", "rb") >>> f <open file u'\u0142\xe4.mp3', mode 'rb' at 0x7f8c25faca08> >>> for i in os.listdir(u'.'): ... if i[-3:] == u"mp3": ... print i ... łä.mp3 and using non-latin chars: >>> f = open(u"चीन.mp3", "w") >>> f <open file u'\u091a\u0940\u0928.mp3', mode 'w' at 0x7f3520079cd8> >>> for i in os.listdir(u"."): ... if i[-3:] == u"mp3": ... print i ... łä.mp3 चीन.mp3 So basicly, one can write an app in python that's using unicode for everyting, including filenames and will be forsed to recode them manually to filesystem encoding for tagpy. It is a bug IMHO. Regards, A. -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org