getttext defaults with unicode

[anglerud]

BPO	563915
Nosy	@malemburg, @loewis, @warsaw

^{Note: these values reflect the state of the issue at the time it was migrated and might not reflect the current state.}

Show more details

GitHub fields:

assignee = 'https://github.com/warsaw'
closed_at = <Date 2002-06-13.20:03:32.000>
created_at = <Date 2002-06-03.15:09:40.000>
labels = ['expert-unicode']
title = 'getttext defaults with unicode'
updated_at = <Date 2002-06-13.20:03:32.000>
user = 'https://bugs.python.org/anglerud'

bugs.python.org fields:

activity = <Date 2002-06-13.20:03:32.000>
actor = 'barry'
assignee = 'barry'
closed = True
closed_date = None
closer = None
components = ['Unicode']
creation = <Date 2002-06-03.15:09:40.000>
creator = 'anglerud'
dependencies = []
files = []
hgrepos = []
issue_num = 563915
keywords = []
message_count = 5.0
messages = ['11041', '11042', '11043', '11044', '11045']
nosy_count = 4.0
nosy_names = ['lemburg', 'loewis', 'barry', 'anglerud']
pr_nums = []
priority = 'normal'
resolution = 'fixed'
stage = None
status = 'closed'
superseder = None
type = None
url = 'https://bugs.python.org/issue563915'
versions = ['Python 2.2']

[anglerud]

When using gettext, and a token is not translated, for
example:
print _('Foo')
the output will be 'Foo', as expected. 
However, if that entry is:
print _(u'Foo')
we get:
Traceback (most recent call last):
  File "./translation.py", line 12, in ?
    print _(u'Foo')
  File "/usr/lib/python2.2/gettext.py", line 195, in
ugettext
    return unicode(tmsg, self._charset)
TypeError: decoding Unicode is not supported

[loewis]

Logged In: YES
user_id=21627

I'm not sure this is a bug: they keys in a message catalog
are really byte strings. So it is an error to use Unicode
strings as arguments to _.

[malemburg]

Logged In: YES
user_id=38388

Barry will know what to do...

[warsaw]

Logged In: YES
user_id=12800

I agree with Martin, you shouldn't be passing unicode
strings to ugettext(). I don't think it's worth doing a
type test in ugettext(); maybe we simply need to improve the
documentation?

[warsaw]

Logged In: YES
user_id=12800

I will shortly post a patch that clarifies the documentation
to describe the GNU gettext recommendation that the source
message string should be us-ascii. As that fixes things in
my mind, I'm closing this bug report.