adding comments to pot files automatically

Question

I want to pull certain comments from my py files that give context to translations, rather than manually editing the .pot file basically i want to go from this python file:

# For Translators: some useful info about the sentence below
_("Some string blah blah")

to this pot file:

# For Translators: some useful info about the sentence below
#: something.py:1
msgid "Some string blah blah"
msgstr ""

Answer 1

I was going to suggest the compiler module, but it ignores comments:

f.py:

# For Translators: some useful info about the sentence below
_("Some string blah blah")

..and the compiler module:

>>> import compiler
>>> m = compiler.parseFile("f.py")
>>> m
Module(None, Stmt([Discard(CallFunc(Name('_'), [Const('Some string blah blah')], None, None))]))

The AST module in Python 2.6 seems to do the same.

Not sure if it's possible, but if you use triple-quoted strings instead..

"""For Translators: some useful info about the sentence below"""
_("Some string blah blah")

..you can reliably parse the Python file with the compiler module:

>>> m = compiler.parseFile("f.py")
>>> m
Module('For Translators: some useful info about the sentence below', Stmt([Discard(CallFunc(Name('_'), [Const('Some string blah blah')], None, None))]))

I made an attempt at writing a mode complete script to extract docstrings - it's incomplete, but seems to grab most docstrings: http://pastie.org/446156 (or on github.com/dbr/so_scripts)

The other, much simpler, option would be to use regular expressions, for example:

f = """# For Translators: some useful info about the sentence below
_("Some string blah blah")
""".split("\n")

import re

for i, line in enumerate(f):
    m = re.findall("\S*# (For Translators: .*)$", line)
    if len(m) > 0 and i != len(f):
        print "Line Number:", i+1
        print "Message:", m
        print "Line:", f[i + 1]

..outputs:

Line Number: 1
Message: ['For Translators: some useful info about the sentence below']
Line: _("Some string blah blah")

Not sure how the .pot file is generated, so I can't be any help at-all with that part..

Answer 2

After much pissing about I found the best way to do this:

#. Translators:
# Blah blah blah
_("String")

Then search for comments with a . like so:

xgettext --language=Python --keyword=_ --add-comments=. --output=test.pot *.py