Special chars and --download

[MatthieuBizien]

Hi

There is an issue with special chars in the path when you download subtitles.

$ ./lm.py ~/Vidéos/ --download eng

2012-02-28 00:28:25,082: LM - ERROR - this path doesnt belong to cash_path /home/matthieu/Vidéos/Phy432_amphi_7.mkv
Traceback (most recent call last):
  File "./lm.py", line 1436, in <module>
    LM.download_subtitle(files, options.download)
  File "./lm.py", line 1005, in download_subtitle
    ref, query = self.download_subtitles_query( files, language )
  File "./lm.py", line 1047, in download_subtitles_query
    filedir = os.path.dirname(f).decode(sys.getdefaultencoding())
UnicodeEncodeError: 'ascii' codec can't encode character u'\xe9' in position 18: ordinal not in range(128)

Regards
Matthieu

[RedRise]

Thank you for this user experience:

Could you please try this same call with debug mode (./lm.py ~/Vidéos/ --download eng --debug).
It will create a lo file in ~/.lm/lm_log.txt
Could you send it to me by mail? polluxxx at gmail dot com

[MatthieuBizien]

Log sent

[RedRise]

This issue should be corrected in last update (tar.gz download also updated)

[MatthieuBizien]

It doesn't work...

[matthieu@matthieu:~/Programmes/lm]$ git pull
Already up-to-date.
[matthieu@matthieu:~/Programmes/lm]$ ./lm.py ~/Vidéos/ --download eng --debug
2012-03-01 04:25:31,896: LM UTIL - INFO - argparse namespace: Namespace(alphabetical=False, confirm=False, debug=True, delete_cache=False, download='eng', files=['/home/matthieu/Vid\xc3\xa9os/'], filter=None, long=False, outline=False, reset=False, reverse=False, show=False, show_imdb=False, upload=False, version=False, very_long=False)
2012-03-01 04:25:31,896: LM UTIL - INFO - arg files type: <type 'str'>
2012-03-01 04:25:31,897: LM UTIL - INFO - file system encoding: UTF-8
2012-03-01 04:25:31,897: LM UTIL - INFO - system encoding: ascii
2012-03-01 04:25:31,897: LM UTIL - INFO - locale encoding: UTF8
2012-03-01 04:25:31,897: LM - INFO - LM initialization
2012-03-01 04:25:31,897: LM - INFO - loading cache_path
2012-03-01 04:25:31,902: LM - INFO - cache_path file loaded successfully
2012-03-01 04:25:31,902: LM - INFO - loading cache_hash
2012-03-01 04:25:31,917: LM - INFO - cache_hash file loaded successfully
2012-03-01 04:25:31,967: LM - INFO - interpreting file/dir argument
2012-03-01 04:25:32,006: LM - INFO - saving caches
2012-03-01 04:25:32,007: LM - INFO - synchronizing caches
2012-03-01 04:25:32,007: LM - INFO - saving cache_path
2012-03-01 04:25:32,013: LM - INFO - saving cache_hash
2012-03-01 04:25:32,029: LM - INFO - cache_hash saved
Traceback (most recent call last):
  File "./lm.py", line 1436, in <module>
    LM.download_subtitle(files, options.download)
  File "./lm.py", line 1005, in download_subtitle
    ref, query = self.download_subtitles_query( files, language )
  File "./lm.py", line 1047, in download_subtitles_query
    filedir = os.path.dirname(f).decode(sys.getdefaultencoding())
UnicodeEncodeError: 'ascii' codec can't encode character u'\xe9' in position 18: ordinal not in range(128)

[RedRise]

I was able to reproduce this bug.
I corrected a string decoding call when download request is built.

I think(/hope?) it will work now.
Thank you for your feedback.

[MatthieuBizien]

It works, thanks :)