deluge|W: unicode mess
[23:36:15] deluge|W: when try to open priv msg in GC for Калоянъ :
Traceback (most recent call last):
File "/home/deluge/CVS/gajim/gajim/src/groupchat_window.py", line 1557, in on_list_treeview_row_activated
gajim.interface.roster.new_chat(u, self.account)
File "/home/deluge/CVS/gajim/gajim/src/roster_window.py", line 1509, in new_chat
gajim.interface.windows[account]['chats']['tabbed'] = \\n File "/home/deluge/CVS/gajim/gajim/src/tabbed_chat_window.py", line 67, in __init__
self.new_tab(contact)
File "/home/deluge/CVS/gajim/gajim/src/tabbed_chat_window.py", line 475, in new_tab
self.restore_conversation(contact.jid)
File "/home/deluge/CVS/gajim/gajim/src/tabbed_chat_window.py", line 889, in restore_conversation
count = gajim.logger.get_no_of_lines(jid)
File "/home/deluge/CVS/gajim/gajim/src/common/logger.py", line 170, in get_no_of_lines
if not os.path.isfile(path_to_file):
File "/usr/lib/python2.4/posixpath.py", line 208, in isfile
st = os.stat(path)
File "/usr/lib/python2.4/encodings/iso8859_2.py", line 18, in encode
return codecs.charmap_encode(input,errors,encoding_map)
UnicodeEncodeError: 'charmap' codec can't encode characters in position 52-58: character maps to \<undefined>
[23:38:25] Калоянъ: hm, deluge|W, what is this 'charmap' codec?
[23:38:45] deluge|W: Калоянъ I have no idee
[23:38:50] Калоянъ: I've never heard about such character codec before
[23:39:32] Калоянъ: oh, may be Gajim is trying to create a log filename with my nickname
[23:40:03] Калоянъ: but the filesystem does not allow this
[23:40:37] Калоянъ: iso-8859-2 does not support any cyrillic characters
[23:41:35] deluge|W: perhaps
[23:42:53] Калоянъ: I suppose this stack trace refers to the line numbers of Gaijm in the trunk, right?
[23:47:02] deluge|W: yes
[23:47:02] Калоянъ: I see
[23:47:02] Калоянъ: let me check this out
[23:47:02] deluge|W: damn, I got very big lag
[23:47:02] Калоянъ: tabbet_chat_window.py:880 contains the line:
[23:47:02] Калоянъ: count = gajim.logger.get_no_of_lines(jid)
[23:47:21] Калоянъ: get_no_of_lines() uses the filesystem, and the jid argument to create a path
[23:47:26] Калоянъ: but currently my JID is:
[23:47:36] Калоянъ: "gajim@conference.jabber.no/Калоянъ"
[23:48:41] Калоянъ: It looks like some characters here are not supported by the host operating system.
[23:49:01] deluge|W: it looks like this
[23:49:05] Калоянъ: ups, I meant _file system_, not operating system
[23:49:29] deluge|W: my file system is JFS
[23:50:04] Калоянъ: I do not know whether JFS supports different file name encodings
[23:50:34] Калоянъ: but may be Gajim have to be prepared to meet file systems capable of US-ASCII only
[23:50:52] deluge|W: It must be
[23:52:08] Калоянъ: When constructing filenames, we Gajim may:
[23:52:11] Калоянъ: (1) get rid of non-US-ASCII characters
[23:52:12] deluge|W: so in this way I created them many work to do
[23:52:26] Калоянъ: (2) escape them somehow (urllib.urlencode(), for example)
[23:53:24] Калоянъ: (a typo above: "we Gajim may" -> "Gajim may")
[23:56:18] Калоянъ: May be we need some sort of filter function to escape filenames, and use it anywhere when a filename is constructed from a JID, or any other parameter that may contain non-US-ASCII characters}