I'd like to parse _('こんにちは').
However pygettext.py doesn't parse _('こんにちは').
pygettext.py said me 'IndexError'.
now I attached pygettext.py.patch to fix a bug.
I show you command history.

$ pygettext.py -o - --verbose konnichiha.py
...
#: konnichiha.py:6
msgid "konnichiha"
msgstr ""

#: konnichiha.py:7
Traceback (most recent call last):
  File "/home/umetaro/local/bin/pygettext.py", line 664, in <module>
    main()
  File "/home/umetaro/local/bin/pygettext.py", line 657, in main
    eater.write(fp)
  File "/home/umetaro/local/bin/pygettext.py", line 497, in write
    print('msgid', normalize(k), file=fp)
  File "/home/umetaro/local/bin/pygettext.py", line 250, in normalize
    s = '"' + escape(s) + '"'
  File "/home/umetaro/local/bin/pygettext.py", line 236, in escape
    s[i] = escapes[ord(s[i])]
IndexError: list index out of range

please use pygettext.py.patch.

$ pygettext.py -o - --verbose konnichiha.py
...
#: konnichiha.py:6
msgid "konnichiha"
msgstr ""

#: konnichiha.py:7
msgid "こんにちは"
msgstr ""

TOO SORRY.

pygettext.py.patch 	umedoblock, 2013-02-08 10:32

is wrong a patch.

please forget it.

Here is a patch for 3.x, which correctly detects input file encoding and correctly escapes non-ascii output files if -E specified (and only if it specified).

For 2.7 we should just negate an argument for make_escapes.

Here is a patch for 2.7. pygettext doesn't try to detect input encoding and transparently works with bytes, but it no longer escapes non-ascii bytes if -E is not specified.

thanks serhiy.storchaka.
I try to use Shift_JIS, UTF-8, ISO-2022-JP and EUC-JP.

your patch detects UTF-8.

However it doesn't detect Shift_JIS, ISO-2022-JP and EUC-JP.
it misunderstand ISO-2022-JP charset is UTF-8.
it raised UnicodeDecodeError when Shift_JIS, EUC-JP.

Please check my test to use konnichiha.sh.

I use just a pygettext_unicode.patch.
don't use a pygettext_unicode-2.7.patch.

Default encoding on Python 3 is UTF-8. You should declare your encoding at the top of file if it differs from UTF-8 or ASCII (i.e. "# -*- coding: euc-jp -*-"). Otherwise Python will reject your file (for Shift_JIS and EUC-JP) or produce incorrect result (for ISO-2022-JP).

$ python3 konnichiha.Shift_JIS.py
  File "konnichiha.Shift_JIS.py", line 5
SyntaxError: Non-UTF-8 code starting with '\x82' in file konnichiha.Shift_JIS.py on line 5, but no encoding declared; see http://python.org/dev/peps/pep-0263/ for details
$ python3 konnichiha.ISO-2022-JP.py
konnichiha
B$3$s$K$A$O

python3 output translate Japanese with pygettext.install().

EVERYTHING IS OK!

please check to use a konnichiha.2.tar.gz.
==============================================
please do below shell command.

$ for f in `find . -name 'konnichiha.*.py'` ; do echo f=$f ; python3 $f ; echo -- ; done
f=./konnichiha.Shift_JIS.py
HELLO ハローで、今日は
日本語をＵＴＦ８にしてコンニチハ
--
f=./konnichiha.UTF-8.py
HELLO ハローで、今日は
日本語をＵＴＦ８にしてコンニチハ
--
f=./konnichiha.ISO-2022-JP.py
HELLO ハローで、今日は
日本語をＵＴＦ８にしてコンニチハ
--
f=./konnichiha.EUC-JP.py
HELLO ハローで、今日は
日本語をＵＴＦ８にしてコンニチハ
--

==============================================
konnichiha script encoding is OK!

$ nkf -g ./konnichiha.*.py                   
./konnichiha.EUC-JP.py: EUC-JP
./konnichiha.ISO-2022-JP.py: ISO-2022-JP
./konnichiha.Shift_JIS.py: Shift_JIS
./konnichiha.UTF-8.py: UTF-8

==============================================
also coding: is OK!
$ head -2 konnichiha.*.py             
==> konnichiha.EUC-JP.py <==
# coding: euc-jp
import gettext

==> konnichiha.ISO-2022-JP.py <==
# coding: iso-2022-jp
import gettext

==> konnichiha.Shift_JIS.py <==
# coding: shift-jis
import gettext

==> konnichiha.UTF-8.py <==
# coding: utf-8
import gettext

==============================================
THANK YOU serhiy.storchaka !

New changeset 49b1fde510a6 by Serhiy Storchaka in branch '2.7':
Issue #17156: pygettext.py now correctly escapes non-ascii characters.
http://hg.python.org/cpython/rev/49b1fde510a6

New changeset cd59b398907d by Serhiy Storchaka in branch '3.2':
Issue #17156: pygettext.py now uses an encoding of source file and correctly
http://hg.python.org/cpython/rev/cd59b398907d

New changeset 062406c06cc1 by Serhiy Storchaka in branch '3.3':
Issue #17156: pygettext.py now uses an encoding of source file and correctly
http://hg.python.org/cpython/rev/062406c06cc1

New changeset 99795d711a40 by Serhiy Storchaka in branch 'default':
Issue #17156: pygettext.py now uses an encoding of source file and correctly
http://hg.python.org/cpython/rev/99795d711a40

Also fixes 19907?