[PATCH 2 of 6] Determine default locale encoding and stdio encoding on start-up

[Marcos Chaves]

Hi Andrey,

if this information helps, bzr crashes on my windows the way it
handles the encodings. For some reason, if I run:

> chcp
Active code page: 0

Then bzr tries to set codepage 'cp0' and crashes. Not sure why, since
locale.getpreferredencoding() gives me "cp1252", and
locale.getdefaultlocale() gives me "('pt_BR', 'cp1252')". Anyway, I
don't know how this works...

There's a patch proposed, but I think it was never applied. See:
http://article.gmane.org/gmane.comp.version-control.bazaar-ng.general/19333

-Marcos

On 11/18/06, Andrey <grooz-work at gorodok.net> wrote:
> I've just looked at how bzr does it. These functions may be of interest for us. :)
>
> Andrey
>
> # taken from http://bazaar-vcs.org/bzr/bzr.dev/bzrlib/osutils.py
> def get_terminal_encoding():
>     """Find the best encoding for printing to the screen.
>
>     This attempts to check both sys.stdout and sys.stdin to see
>     what encoding they are in, and if that fails it falls back to
>     bzrlib.user_encoding.
>     The problem is that on Windows, locale.getpreferredencoding()
>     is not the same encoding as that used by the console:
>     http://mail.python.org/pipermail/python-list/2003-May/162357.html
>
>     On my standard US Windows XP, the preferred encoding is
>     cp1252, but the console is cp437
>     """
>     output_encoding = getattr(sys.stdout, 'encoding', None)
>     if not output_encoding:
>         input_encoding = getattr(sys.stdin, 'encoding', None)
>         if not input_encoding:
>             output_encoding = bzrlib.user_encoding
>             mutter('encoding stdout as bzrlib.user_encoding %r', output_encoding)
>         else:
>             output_encoding = input_encoding
>             mutter('encoding stdout as sys.stdin encoding %r', output_encoding)
>     else:
>         mutter('encoding stdout as sys.stdout encoding %r', output_encoding)
>     return output_encoding
>
> def get_user_encoding():
>     """Find out what the preferred user encoding is.
>
>     This is generally the encoding that is used for command line parameters
>     and file contents. This may be different from the terminal encoding
>     or the filesystem encoding.
>
>     :return: A string defining the preferred user encoding
>     """
>     global _cached_user_encoding
>     if _cached_user_encoding is not None:
>         return _cached_user_encoding
>
>     if sys.platform == 'darwin':
>         # work around egregious python 2.4 bug
>         sys.platform = 'posix'
>         try:
>             import locale
>         finally:
>             sys.platform = 'darwin'
>     else:
>         import locale
>
>     try:
>         _cached_user_encoding = locale.getpreferredencoding()
>     except locale.Error, e:
>         sys.stderr.write('bzr: warning: %s\n'
>                          '  Could not determine what text encoding to use.\n'
>                          '  This error usually means your Python interpreter\n'
>                          '  doesn\'t support the locale set by $LANG (%s)\n'
>                          "  Continuing with ascii encoding.\n"
>                          % (e, os.environ.get('LANG')))
>
>     if _cached_user_encoding is None:
>         _cached_user_encoding = 'ascii'
>     return _cached_user_encoding
> _______________________________________________
> Mercurial-devel mailing list
> Mercurial-devel at selenic.com
> http://selenic.com/mailman/listinfo/mercurial-devel
>