def test_sorted_po(self):
catalog = Catalog()
catalog.add(
u('bar'),
locations=[('utils.py', 3)],
user_comments=['Comment About `bar` with', 'multiple lines.'])
catalog.add((u('foo'), u('foos')), (u('Voh'), u('Voeh')),
locations=[('main.py', 1)])
buf = BytesIO()
pofile.write_po(buf, catalog, sort_output=True)
value = buf.getvalue().strip()
assert b('''\
# Comment About `bar` with
# multiple lines.
#: utils.py:3
msgid "bar"
msgstr ""
#: main.py:1
msgid "foo"
msgid_plural "foos"
msgstr[0] "Voh"
msgstr[1] "Voeh"''') in value
assert value.find(b('msgid ""')) < value.find(
b('msgid "bar"')) < value.find(b('msgid "foo"'))
def test_no_wrap_and_width_behaviour_on_comments(self):
catalog = Catalog()
catalog.add(
"Pretty dam long message id, which must really be big "
"to test this wrap behaviour, if not it won't work.",
locations=[("fake.py", n) for n in range(1, 30)])
buf = BytesIO()
pofile.write_po(buf, catalog, width=None, omit_header=True)
self.assertEqual(
b("""\
#: fake.py:1 fake.py:2 fake.py:3 fake.py:4 fake.py:5 fake.py:6 fake.py:7
#: fake.py:8 fake.py:9 fake.py:10 fake.py:11 fake.py:12 fake.py:13 fake.py:14
#: fake.py:15 fake.py:16 fake.py:17 fake.py:18 fake.py:19 fake.py:20 fake.py:21
#: fake.py:22 fake.py:23 fake.py:24 fake.py:25 fake.py:26 fake.py:27 fake.py:28
#: fake.py:29
msgid "pretty dam long message id, which must really be big to test this wrap behaviour, if not it won't work."
msgstr ""
"""),
buf.getvalue().lower())
buf = BytesIO()
pofile.write_po(buf, catalog, width=100, omit_header=True)
self.assertEqual(
b("""\
#: fake.py:1 fake.py:2 fake.py:3 fake.py:4 fake.py:5 fake.py:6 fake.py:7 fake.py:8 fake.py:9 fake.py:10
#: fake.py:11 fake.py:12 fake.py:13 fake.py:14 fake.py:15 fake.py:16 fake.py:17 fake.py:18 fake.py:19
#: fake.py:20 fake.py:21 fake.py:22 fake.py:23 fake.py:24 fake.py:25 fake.py:26 fake.py:27 fake.py:28
#: fake.py:29
msgid ""
"pretty dam long message id, which must really be big to test this wrap behaviour, if not it won't"
" work."
msgstr ""
"""),
buf.getvalue().lower())
def test_no_wrap_and_width_behaviour_on_comments(self):
catalog = Catalog()
catalog.add("Pretty dam long message id, which must really be big "
"to test this wrap behaviour, if not it won't work.",
locations=[("fake.py", n) for n in range(1, 30)])
buf = BytesIO()
pofile.write_po(buf, catalog, width=None, omit_header=True)
self.assertEqual(b("""\
#: fake.py:1 fake.py:2 fake.py:3 fake.py:4 fake.py:5 fake.py:6 fake.py:7
#: fake.py:8 fake.py:9 fake.py:10 fake.py:11 fake.py:12 fake.py:13 fake.py:14
#: fake.py:15 fake.py:16 fake.py:17 fake.py:18 fake.py:19 fake.py:20 fake.py:21
#: fake.py:22 fake.py:23 fake.py:24 fake.py:25 fake.py:26 fake.py:27 fake.py:28
#: fake.py:29
msgid "pretty dam long message id, which must really be big to test this wrap behaviour, if not it won't work."
msgstr ""
"""), buf.getvalue().lower())
buf = BytesIO()
pofile.write_po(buf, catalog, width=100, omit_header=True)
self.assertEqual(b("""\
#: fake.py:1 fake.py:2 fake.py:3 fake.py:4 fake.py:5 fake.py:6 fake.py:7 fake.py:8 fake.py:9 fake.py:10
#: fake.py:11 fake.py:12 fake.py:13 fake.py:14 fake.py:15 fake.py:16 fake.py:17 fake.py:18 fake.py:19
#: fake.py:20 fake.py:21 fake.py:22 fake.py:23 fake.py:24 fake.py:25 fake.py:26 fake.py:27 fake.py:28
#: fake.py:29
msgid ""
"pretty dam long message id, which must really be big to test this wrap behaviour, if not it won't"
" work."
msgstr ""
"""), buf.getvalue().lower())
def test_npgettext(self):
self.assertEqualTypeToo('Voh1',
self.translations.ngettext('foo1', 'foos1', 1))
self.assertEqualTypeToo('Vohs1',
self.translations.ngettext('foo1', 'foos1', 2))
self.assertEqualTypeToo(
b('VohCTX1'), self.translations.npgettext('foo', 'foo1', 'foos1',
1))
self.assertEqualTypeToo(
b('VohsCTX1'),
self.translations.npgettext('foo', 'foo1', 'foos1', 2))
def test_ldnpgettext(self):
self.assertEqualTypeToo(
b('VohD1'), self.translations.ldngettext('messages1', 'foo1', 'foos1', 1))
self.assertEqualTypeToo(
b('VohsD1'), self.translations.ldngettext('messages1', 'foo1', 'foos1', 2))
self.assertEqualTypeToo(
b('VohCTXD1'), self.translations.ldnpgettext('messages1', 'foo', 'foo1',
'foos1', 1))
self.assertEqualTypeToo(
b('VohsCTXD1'), self.translations.ldnpgettext('messages1', 'foo', 'foo1',
'foos1', 2))
def test_lnpgettext(self):
self.assertEqualTypeToo(b('Voh1'),
self.translations.lngettext('foo1', 'foos1', 1))
self.assertEqualTypeToo(b('Vohs1'),
self.translations.lngettext('foo1', 'foos1', 2))
self.assertEqualTypeToo(b('VohCTX1'),
self.translations.lnpgettext('foo', 'foo1',
'foos1', 1))
self.assertEqualTypeToo(b('VohsCTX1'),
self.translations.lnpgettext('foo', 'foo1',
'foos1', 2))
def test_ldnpgettext(self):
self.assertEqualTypeToo(
b('VohD1'),
self.translations.ldngettext('messages1', 'foo1', 'foos1', 1))
self.assertEqualTypeToo(
b('VohsD1'),
self.translations.ldngettext('messages1', 'foo1', 'foos1', 2))
self.assertEqualTypeToo(
b('VohCTXD1'),
self.translations.ldnpgettext('messages1', 'foo', 'foo1', 'foos1',
1))
self.assertEqualTypeToo(
b('VohsCTXD1'),
self.translations.ldnpgettext('messages1', 'foo', 'foo1', 'foos1',
2))
def test_wrap_long_lines_in_header(self):
"""
Verify that long lines in the header comment are wrapped correctly.
"""
catalog = Catalog(project='AReallyReallyLongNameForAProject',
revision_date=datetime(2007, 4, 1))
buf = BytesIO()
pofile.write_po(buf, catalog)
self.assertEqual(b('''\
# Translations template for AReallyReallyLongNameForAProject.
# Copyright (C) 2007 ORGANIZATION
# This file is distributed under the same license as the
# AReallyReallyLongNameForAProject project.
# FIRST AUTHOR <EMAIL@ADDRESS>, 2007.
#
#, fuzzy'''), b('\n').join(buf.getvalue().splitlines()[:7]))
def test_with_context(self):
buf = StringIO(r'''# Some string in the menu
#: main.py:1
msgctxt "Menu"
msgid "foo"
msgstr "Voh"
# Another string in the menu
#: main.py:2
msgctxt "Menu"
msgid "bar"
msgstr "Bahr"
''')
catalog = pofile.read_po(buf, ignore_obsolete=True)
self.assertEqual(2, len(catalog))
message = catalog.get('foo', context='Menu')
self.assertEqual('Menu', message.context)
message = catalog.get('bar', context='Menu')
self.assertEqual('Menu', message.context)
# And verify it pass through write_po
out_buf = BytesIO()
pofile.write_po(out_buf, catalog, omit_header=True)
assert out_buf.getvalue().strip() == b(buf.getvalue().strip()), \
out_buf.getvalue()
def test_with_context(self):
buf = StringIO(r'''# Some string in the menu
#: main.py:1
msgctxt "Menu"
msgid "foo"
msgstr "Voh"
# Another string in the menu
#: main.py:2
msgctxt "Menu"
msgid "bar"
msgstr "Bahr"
''')
catalog = pofile.read_po(buf, ignore_obsolete=True)
self.assertEqual(2, len(catalog))
message = catalog.get('foo', context='Menu')
self.assertEqual('Menu', message.context)
message = catalog.get('bar', context='Menu')
self.assertEqual('Menu', message.context)
# And verify it pass through write_po
out_buf = BytesIO()
pofile.write_po(out_buf, catalog, omit_header=True)
assert out_buf.getvalue().strip() == b(buf.getvalue().strip()), \
out_buf.getvalue()
def test_wrap_long_lines(self):
text = """Here's some text where
white space and line breaks matter, and should
not be removed
"""
catalog = Catalog()
catalog.add(text, locations=[('main.py', 1)])
buf = BytesIO()
pofile.write_po(buf,
catalog,
no_location=True,
omit_header=True,
width=42)
self.assertEqual(
b(r'''msgid ""
"Here's some text where \n"
"white space and line breaks matter, and"
" should\n"
"\n"
"not be removed\n"
"\n"
msgstr ""'''),
buf.getvalue().strip())
def test_po_with_multiline_obsolete_message(self):
catalog = Catalog()
catalog.add(u('foo'), u('Voh'), locations=[('main.py', 1)])
msgid = r"""Here's a message that covers
multiple lines, and should still be handled
correctly.
"""
msgstr = r"""Here's a message that covers
multiple lines, and should still be handled
correctly.
"""
catalog.obsolete[msgid] = Message(msgid,
msgstr,
locations=[('utils.py', 3)])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True)
self.assertEqual(
b(r'''#: main.py:1
msgid "foo"
msgstr "Voh"
#~ msgid ""
#~ "Here's a message that covers\n"
#~ "multiple lines, and should still be handled\n"
#~ "correctly.\n"
#~ msgstr ""
#~ "Here's a message that covers\n"
#~ "multiple lines, and should still be handled\n"
#~ "correctly.\n"'''),
buf.getvalue().strip())
def test_po_with_multiline_obsolete_message(self):
catalog = Catalog()
catalog.add(u('foo'), u('Voh'), locations=[('main.py', 1)])
msgid = r"""Here's a message that covers
multiple lines, and should still be handled
correctly.
"""
msgstr = r"""Here's a message that covers
multiple lines, and should still be handled
correctly.
"""
catalog.obsolete[msgid] = Message(msgid, msgstr,
locations=[('utils.py', 3)])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True)
self.assertEqual(b(r'''#: main.py:1
msgid "foo"
msgstr "Voh"
#~ msgid ""
#~ "Here's a message that covers\n"
#~ "multiple lines, and should still be handled\n"
#~ "correctly.\n"
#~ msgstr ""
#~ "Here's a message that covers\n"
#~ "multiple lines, and should still be handled\n"
#~ "correctly.\n"'''), buf.getvalue().strip())
def test_write_po_file_with_specified_charset(self):
catalog = Catalog(charset='iso-8859-1')
catalog.add('foo', u('\xe4\xf6\xfc'), locations=[('main.py', 1)])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=False)
po_file = buf.getvalue().strip()
assert b(r'"Content-Type: text/plain; charset=iso-8859-1\n"') in po_file
assert u('msgstr "\xe4\xf6\xfc"').encode('iso-8859-1') in po_file
def test_write_po_file_with_specified_charset(self):
catalog = Catalog(charset='iso-8859-1')
catalog.add('foo', u('\xe4\xf6\xfc'), locations=[('main.py', 1)])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=False)
po_file = buf.getvalue().strip()
assert b(
r'"Content-Type: text/plain; charset=iso-8859-1\n"') in po_file
assert u('msgstr "\xe4\xf6\xfc"').encode('iso-8859-1') in po_file
def test_join_locations(self):
catalog = Catalog()
catalog.add(u('foo'), locations=[('main.py', 1)])
catalog.add(u('foo'), locations=[('utils.py', 3)])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True)
self.assertEqual(b('''#: main.py:1 utils.py:3
msgid "foo"
msgstr ""'''), buf.getvalue().strip())
def test_wrap_long_lines_in_header(self):
"""
Verify that long lines in the header comment are wrapped correctly.
"""
catalog = Catalog(project='AReallyReallyLongNameForAProject',
revision_date=datetime(2007, 4, 1))
buf = BytesIO()
pofile.write_po(buf, catalog)
self.assertEqual(
b('''\
# Translations template for AReallyReallyLongNameForAProject.
# Copyright (C) 2007 ORGANIZATION
# This file is distributed under the same license as the
# AReallyReallyLongNameForAProject project.
# FIRST AUTHOR <EMAIL@ADDRESS>, 2007.
#
#, fuzzy'''),
b('\n').join(buf.getvalue().splitlines()[:7]))
def test_duplicate_comments(self):
catalog = Catalog()
catalog.add(u('foo'), auto_comments=['A comment'])
catalog.add(u('foo'), auto_comments=['A comment'])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True)
self.assertEqual(b('''#. A comment
msgid "foo"
msgstr ""'''), buf.getvalue().strip())
def test_po_with_previous_msgid(self):
catalog = Catalog()
catalog.add(u('foo'), u('Voh'), locations=[('main.py', 1)],
previous_id=u('fo'))
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True, include_previous=True)
self.assertEqual(b('''#: main.py:1
#| msgid "fo"
msgid "foo"
msgstr "Voh"'''), buf.getvalue().strip())
def test_duplicate_comments(self):
catalog = Catalog()
catalog.add(u('foo'), auto_comments=['A comment'])
catalog.add(u('foo'), auto_comments=['A comment'])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True)
self.assertEqual(b('''#. A comment
msgid "foo"
msgstr ""'''),
buf.getvalue().strip())
def test_po_with_obsolete_message_ignored(self):
catalog = Catalog()
catalog.add(u('foo'), u('Voh'), locations=[('main.py', 1)])
catalog.obsolete['bar'] = Message(u('bar'), u('Bahr'),
locations=[('utils.py', 3)],
user_comments=['User comment'])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True, ignore_obsolete=True)
self.assertEqual(b('''#: main.py:1
msgid "foo"
msgstr "Voh"'''), buf.getvalue().strip())
def test_join_locations(self):
catalog = Catalog()
catalog.add(u('foo'), locations=[('main.py', 1)])
catalog.add(u('foo'), locations=[('utils.py', 3)])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True)
self.assertEqual(
b('''#: main.py:1 utils.py:3
msgid "foo"
msgstr ""'''),
buf.getvalue().strip())
def test_po_with_obsolete_message_ignored(self):
catalog = Catalog()
catalog.add(u('foo'), u('Voh'), locations=[('main.py', 1)])
catalog.obsolete['bar'] = Message(u('bar'),
u('Bahr'),
locations=[('utils.py', 3)],
user_comments=['User comment'])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True, ignore_obsolete=True)
self.assertEqual(b('''#: main.py:1
msgid "foo"
msgstr "Voh"'''),
buf.getvalue().strip())
def test_po_with_previous_msgid(self):
catalog = Catalog()
catalog.add(u('foo'),
u('Voh'),
locations=[('main.py', 1)],
previous_id=u('fo'))
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True, include_previous=True)
self.assertEqual(
b('''#: main.py:1
#| msgid "fo"
msgid "foo"
msgstr "Voh"'''),
buf.getvalue().strip())
def test_sorted_po(self):
catalog = Catalog()
catalog.add(u('bar'), locations=[('utils.py', 3)],
user_comments=['Comment About `bar` with',
'multiple lines.'])
catalog.add((u('foo'), u('foos')), (u('Voh'), u('Voeh')),
locations=[('main.py', 1)])
buf = BytesIO()
pofile.write_po(buf, catalog, sort_output=True)
value = buf.getvalue().strip()
assert b('''\
# Comment About `bar` with
# multiple lines.
#: utils.py:3
msgid "bar"
msgstr ""
#: main.py:1
msgid "foo"
msgid_plural "foos"
msgstr[0] "Voh"
msgstr[1] "Voeh"''') in value
assert value.find(b('msgid ""')) < value.find(b('msgid "bar"')) < value.find(b('msgid "foo"'))
def test_wrap_locations_with_hyphens(self):
catalog = Catalog()
catalog.add(u('foo'), locations=[
('doupy/templates/base/navmenu.inc.html.py', 60)
])
catalog.add(u('foo'), locations=[
('doupy/templates/job-offers/helpers.html', 22)
])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True)
self.assertEqual(b('''#: doupy/templates/base/navmenu.inc.html.py:60
#: doupy/templates/job-offers/helpers.html:22
msgid "foo"
msgstr ""'''), buf.getvalue().strip())
def test_wrap_long_lines_with_long_word(self):
text = """Here's some text that
includesareallylongwordthatmightbutshouldnt throw us into an infinite loop
"""
catalog = Catalog()
catalog.add(text, locations=[('main.py', 1)])
buf = BytesIO()
pofile.write_po(buf, catalog, no_location=True, omit_header=True,
width=32)
self.assertEqual(b(r'''msgid ""
"Here's some text that\n"
"includesareallylongwordthatmightbutshouldnt"
" throw us into an infinite "
"loop\n"
msgstr ""'''), buf.getvalue().strip())
def test_wrap_locations_with_hyphens(self):
catalog = Catalog()
catalog.add(u('foo'),
locations=[('doupy/templates/base/navmenu.inc.html.py', 60)
])
catalog.add(u('foo'),
locations=[('doupy/templates/job-offers/helpers.html', 22)
])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True)
self.assertEqual(
b('''#: doupy/templates/base/navmenu.inc.html.py:60
#: doupy/templates/job-offers/helpers.html:22
msgid "foo"
msgstr ""'''),
buf.getvalue().strip())
def test_pot_with_translator_comments(self):
catalog = Catalog()
catalog.add(u('foo'), locations=[('main.py', 1)],
auto_comments=['Comment About `foo`'])
catalog.add(u('bar'), locations=[('utils.py', 3)],
user_comments=['Comment About `bar` with',
'multiple lines.'])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True)
self.assertEqual(b('''#. Comment About `foo`
#: main.py:1
msgid "foo"
msgstr ""
# Comment About `bar` with
# multiple lines.
#: utils.py:3
msgid "bar"
msgstr ""'''), buf.getvalue().strip())
def test_wrap_long_lines_with_long_word(self):
text = """Here's some text that
includesareallylongwordthatmightbutshouldnt throw us into an infinite loop
"""
catalog = Catalog()
catalog.add(text, locations=[('main.py', 1)])
buf = BytesIO()
pofile.write_po(buf,
catalog,
no_location=True,
omit_header=True,
width=32)
self.assertEqual(
b(r'''msgid ""
"Here's some text that\n"
"includesareallylongwordthatmightbutshouldnt"
" throw us into an infinite "
"loop\n"
msgstr ""'''),
buf.getvalue().strip())
def test_wrap_long_lines(self):
text = """Here's some text where
white space and line breaks matter, and should
not be removed
"""
catalog = Catalog()
catalog.add(text, locations=[('main.py', 1)])
buf = BytesIO()
pofile.write_po(buf, catalog, no_location=True, omit_header=True,
width=42)
self.assertEqual(b(r'''msgid ""
"Here's some text where \n"
"white space and line breaks matter, and"
" should\n"
"\n"
"not be removed\n"
"\n"
msgstr ""'''), buf.getvalue().strip())
def test_pot_with_translator_comments(self):
catalog = Catalog()
catalog.add(u('foo'),
locations=[('main.py', 1)],
auto_comments=['Comment About `foo`'])
catalog.add(
u('bar'),
locations=[('utils.py', 3)],
user_comments=['Comment About `bar` with', 'multiple lines.'])
buf = BytesIO()
pofile.write_po(buf, catalog, omit_header=True)
self.assertEqual(
b('''#. Comment About `foo`
#: main.py:1
msgid "foo"
msgstr ""
# Comment About `bar` with
# multiple lines.
#: utils.py:3
msgid "bar"
msgstr ""'''),
buf.getvalue().strip())
def read_mo(fileobj):
"""Read a binary MO file from the given file-like object and return a
corresponding `Catalog` object.
:param fileobj: the file-like object to read the MO file from
:return: a catalog object representing the parsed MO file
:rtype: `Catalog`
:note: The implementation of this function is heavily based on the
``GNUTranslations._parse`` method of the ``gettext`` module in the
standard library.
"""
catalog = Catalog()
headers = {}
filename = getattr(fileobj, 'name', '')
buf = fileobj.read()
buflen = len(buf)
unpack = struct.unpack
# Parse the .mo file header, which consists of 5 little endian 32
# bit words.
magic = unpack('<I', buf[:4])[0] # Are we big endian or little endian?
if magic == LE_MAGIC:
version, msgcount, origidx, transidx = unpack('<4I', buf[4:20])
ii = '<II'
elif magic == BE_MAGIC:
version, msgcount, origidx, transidx = unpack('>4I', buf[4:20])
ii = '>II'
else:
raise IOError(0, 'Bad magic number', filename)
# Now put all messages from the .mo file buffer into the catalog
# dictionary
for i in xrange(0, msgcount):
mlen, moff = unpack(ii, buf[origidx:origidx + 8])
mend = moff + mlen
tlen, toff = unpack(ii, buf[transidx:transidx + 8])
tend = toff + tlen
if mend < buflen and tend < buflen:
msg = buf[moff:mend]
tmsg = buf[toff:tend]
else:
raise IOError(0, 'File is corrupt', filename)
# See if we're looking at GNU .mo conventions for metadata
if mlen == 0:
# Catalog description
lastkey = key = None
for item in tmsg.splitlines():
item = item.strip()
if not item:
continue
if b(':') in item:
key, value = item.split(b(':'), 1)
lastkey = key = key.strip().lower()
headers[key] = value.strip()
elif lastkey:
headers[lastkey] += '\n' + item
if b('\x04') in msg: # context
ctxt, msg = msg.split(b('\x04'))
else:
ctxt = None
if b('\x00') in msg: # plural forms
msg = msg.split(b('\x00'))
tmsg = tmsg.split(b('\x00'))
if catalog.charset:
msg = [x.decode(catalog.charset) for x in msg]
tmsg = [x.decode(catalog.charset) for x in tmsg]
else:
if catalog.charset:
msg = msg.decode(catalog.charset)
tmsg = tmsg.decode(catalog.charset)
catalog[msg] = Message(msg, tmsg, context=ctxt)
# advance to next entry in the seek tables
origidx += 8
transidx += 8
catalog.mime_headers = list(headers.items())
return catalog
def test_ldpgettext(self):
self.assertEqualTypeToo(
b('VohD'), self.translations.ldgettext('messages1', 'foo'))
self.assertEqualTypeToo(
b('VohCTXD'),
self.translations.ldpgettext('messages1', 'foo', 'foo'))
def test_lpgettext(self):
self.assertEqualTypeToo(b('Voh'), self.translations.lgettext('foo'))
self.assertEqualTypeToo(b('VohCTX'),
self.translations.lpgettext('foo', 'foo'))
>>> print(list(distinct('foobar')))
['f', 'o', 'b', 'a', 'r']
:param iterable: the iterable collection providing the data
:return: the distinct items in the collection
:rtype: ``iterator``
"""
seen = set()
for item in iter(iterable):
if item not in seen:
yield item
seen.add(item)
# Regexp to match python magic encoding line
PYTHON_MAGIC_COMMENT_re = re.compile(
b(r'[ \t\f]* \# .* coding[=:][ \t]*([-\w.]+)'), re.VERBOSE)
def parse_encoding(fp):
"""Deduce the encoding of a source file from magic comment.
It does this in the same way as the `Python interpreter`__
.. __: http://docs.python.org/ref/encodings.html
The ``fp`` argument should be a seekable file object.
(From Jeff Dairiki)
"""
pos = fp.tell()
fp.seek(0)
try:
line1 = fp.readline()
def read_mo(fileobj):
"""Read a binary MO file from the given file-like object and return a
corresponding `Catalog` object.
:param fileobj: the file-like object to read the MO file from
:return: a catalog object representing the parsed MO file
:rtype: `Catalog`
:note: The implementation of this function is heavily based on the
``GNUTranslations._parse`` method of the ``gettext`` module in the
standard library.
"""
catalog = Catalog()
headers = {}
filename = getattr(fileobj, 'name', '')
buf = fileobj.read()
buflen = len(buf)
unpack = struct.unpack
# Parse the .mo file header, which consists of 5 little endian 32
# bit words.
magic = unpack('<I', buf[:4])[0] # Are we big endian or little endian?
if magic == LE_MAGIC:
version, msgcount, origidx, transidx = unpack('<4I', buf[4:20])
ii = '<II'
elif magic == BE_MAGIC:
version, msgcount, origidx, transidx = unpack('>4I', buf[4:20])
ii = '>II'
else:
raise IOError(0, 'Bad magic number', filename)
# Now put all messages from the .mo file buffer into the catalog
# dictionary
for i in xrange(0, msgcount):
mlen, moff = unpack(ii, buf[origidx:origidx + 8])
mend = moff + mlen
tlen, toff = unpack(ii, buf[transidx:transidx + 8])
tend = toff + tlen
if mend < buflen and tend < buflen:
msg = buf[moff:mend]
tmsg = buf[toff:tend]
else:
raise IOError(0, 'File is corrupt', filename)
# See if we're looking at GNU .mo conventions for metadata
if mlen == 0:
# Catalog description
lastkey = key = None
for item in tmsg.splitlines():
item = item.strip()
if not item:
continue
if b(':') in item:
key, value = item.split(b(':'), 1)
lastkey = key = key.strip().lower()
headers[key] = value.strip()
elif lastkey:
headers[lastkey] += '\n' + item
if b('\x04') in msg: # context
ctxt, msg = msg.split(b('\x04'))
else:
ctxt = None
if b('\x00') in msg: # plural forms
msg = msg.split(b('\x00'))
tmsg = tmsg.split(b('\x00'))
if catalog.charset:
msg = [x.decode(catalog.charset) for x in msg]
tmsg = [x.decode(catalog.charset) for x in tmsg]
else:
if catalog.charset:
msg = msg.decode(catalog.charset)
tmsg = tmsg.decode(catalog.charset)
catalog[msg] = Message(msg, tmsg, context=ctxt)
# advance to next entry in the seek tables
origidx += 8
transidx += 8
catalog.mime_headers = list(headers.items())
return catalog
def test_lpgettext(self):
self.assertEqualTypeToo(b('Voh'), self.translations.lgettext('foo'))
self.assertEqualTypeToo(b('VohCTX'), self.translations.lpgettext('foo',
'foo'))
def write_mo(fileobj, catalog, use_fuzzy=False):
"""Write a catalog to the specified file-like object using the GNU MO file
format.
>>> from babel.messages import Catalog
>>> from babel.compat import BytesIO, GNUTranslations
>>> catalog = Catalog(locale='en_US')
>>> catalog.add('foo', 'Voh')
<Message ...>
>>> catalog.add((u('bar'), u('baz')), (u('Bahr'), u('Batz')))
<Message ...>
>>> catalog.add('fuz', 'Futz', flags=['fuzzy'])
<Message ...>
>>> catalog.add('Fizz', '')
<Message ...>
>>> catalog.add(('Fuzz', 'Fuzzes'), ('', ''))
<Message ...>
>>> buf = BytesIO()
>>> write_mo(buf, catalog)
>>> _ = buf.seek(0)
>>> translations = GNUTranslations(fp=buf)
>>> translations.ugettext('foo') == u('Voh')
True
>>> translations.ungettext('bar', 'baz', 1) == u('Bahr')
True
>>> translations.ungettext('bar', 'baz', 2) == u('Batz')
True
>>> translations.ugettext('fuz') == u('fuz')
True
>>> translations.ugettext('Fizz') == u('Fizz')
True
>>> translations.ugettext('Fuzz') == u('Fuzz')
True
>>> translations.ugettext('Fuzzes') == u('Fuzzes')
True
:param fileobj: the file-like object to write to
:param catalog: the `Catalog` instance
:param use_fuzzy: whether translations marked as "fuzzy" should be included
in the output
"""
messages = list(catalog)
if not use_fuzzy:
messages[1:] = [m for m in messages[1:] if not m.fuzzy]
messages.sort()
ids = strs = b('')
offsets = []
for message in messages:
# For each string, we need size and file offset. Each string is NUL
# terminated; the NUL does not count into the size.
if message.pluralizable:
msgid = b('\x00').join(
[msgid.encode(catalog.charset) for msgid in message.id])
msgstrs = []
for idx, string in enumerate(message.string):
if not string:
msgstrs.append(message.id[min(int(idx), 1)])
else:
msgstrs.append(string)
msgstr = b('\x00').join(
[msgstr.encode(catalog.charset) for msgstr in msgstrs])
else:
msgid = message.id.encode(catalog.charset)
if not message.string:
msgstr = message.id.encode(catalog.charset)
else:
msgstr = message.string.encode(catalog.charset)
if message.context:
msgid = b('\x04').join(
[message.context.encode(catalog.charset), msgid])
offsets.append((len(ids), len(msgid), len(strs), len(msgstr)))
ids += msgid + b('\x00')
strs += msgstr + b('\x00')
# The header is 7 32-bit unsigned integers. We don't use hash tables, so
# the keys start right after the index tables.
keystart = 7 * 4 + 16 * len(messages)
valuestart = keystart + len(ids)
# The string table first has the list of keys, then the list of values.
# Each entry has first the size of the string, then the file offset.
koffsets = []
voffsets = []
for o1, l1, o2, l2 in offsets:
koffsets += [l1, o1 + keystart]
voffsets += [l2, o2 + valuestart]
offsets = koffsets + voffsets
fileobj.write(
struct.pack(
'Iiiiiii',
LE_MAGIC, # magic
0, # version
len(messages), # number of entries
7 * 4, # start of key index
7 * 4 + len(messages) * 8, # start of value index
0,
0 # size and offset of hash table
))
if PY3:
fileobj.write(array.array("i", offsets).tobytes())
else:
fileobj.write(array.array("i", offsets).tostring())
fileobj.write(ids + strs)
def test_ldpgettext(self):
self.assertEqualTypeToo(
b('VohD'), self.translations.ldgettext('messages1', 'foo'))
self.assertEqualTypeToo(
b('VohCTXD'), self.translations.ldpgettext('messages1', 'foo', 'foo'))
>>> print(list(distinct('foobar')))
['f', 'o', 'b', 'a', 'r']
:param iterable: the iterable collection providing the data
:return: the distinct items in the collection
:rtype: ``iterator``
"""
seen = set()
for item in iter(iterable):
if item not in seen:
yield item
seen.add(item)
# Regexp to match python magic encoding line
PYTHON_MAGIC_COMMENT_re = re.compile(b(r"[ \t\f]* \# .* coding[=:][ \t]*([-\w.]+)"), re.VERBOSE)
def parse_encoding(fp):
"""Deduce the encoding of a source file from magic comment.
It does this in the same way as the `Python interpreter`__
.. __: http://docs.python.org/ref/encodings.html
The ``fp`` argument should be a seekable file object.
(From Jeff Dairiki)
"""
pos = fp.tell()
fp.seek(0)
def write_mo(fileobj, catalog, use_fuzzy=False):
"""Write a catalog to the specified file-like object using the GNU MO file
format.
>>> from babel.messages import Catalog
>>> from babel.compat import BytesIO, GNUTranslations
>>> catalog = Catalog(locale='en_US')
>>> catalog.add('foo', 'Voh')
<Message ...>
>>> catalog.add((u('bar'), u('baz')), (u('Bahr'), u('Batz')))
<Message ...>
>>> catalog.add('fuz', 'Futz', flags=['fuzzy'])
<Message ...>
>>> catalog.add('Fizz', '')
<Message ...>
>>> catalog.add(('Fuzz', 'Fuzzes'), ('', ''))
<Message ...>
>>> buf = BytesIO()
>>> write_mo(buf, catalog)
>>> _ = buf.seek(0)
>>> translations = GNUTranslations(fp=buf)
>>> translations.ugettext('foo') == u('Voh')
True
>>> translations.ungettext('bar', 'baz', 1) == u('Bahr')
True
>>> translations.ungettext('bar', 'baz', 2) == u('Batz')
True
>>> translations.ugettext('fuz') == u('fuz')
True
>>> translations.ugettext('Fizz') == u('Fizz')
True
>>> translations.ugettext('Fuzz') == u('Fuzz')
True
>>> translations.ugettext('Fuzzes') == u('Fuzzes')
True
:param fileobj: the file-like object to write to
:param catalog: the `Catalog` instance
:param use_fuzzy: whether translations marked as "fuzzy" should be included
in the output
"""
messages = list(catalog)
if not use_fuzzy:
messages[1:] = [m for m in messages[1:] if not m.fuzzy]
messages.sort()
ids = strs = b('')
offsets = []
for message in messages:
# For each string, we need size and file offset. Each string is NUL
# terminated; the NUL does not count into the size.
if message.pluralizable:
msgid = b('\x00').join([
msgid.encode(catalog.charset) for msgid in message.id
])
msgstrs = []
for idx, string in enumerate(message.string):
if not string:
msgstrs.append(message.id[min(int(idx), 1)])
else:
msgstrs.append(string)
msgstr = b('\x00').join([
msgstr.encode(catalog.charset) for msgstr in msgstrs
])
else:
msgid = message.id.encode(catalog.charset)
if not message.string:
msgstr = message.id.encode(catalog.charset)
else:
msgstr = message.string.encode(catalog.charset)
if message.context:
msgid = b('\x04').join([message.context.encode(catalog.charset),
msgid])
offsets.append((len(ids), len(msgid), len(strs), len(msgstr)))
ids += msgid + b('\x00')
strs += msgstr + b('\x00')
# The header is 7 32-bit unsigned integers. We don't use hash tables, so
# the keys start right after the index tables.
keystart = 7 * 4 + 16 * len(messages)
valuestart = keystart + len(ids)
# The string table first has the list of keys, then the list of values.
# Each entry has first the size of the string, then the file offset.
koffsets = []
voffsets = []
for o1, l1, o2, l2 in offsets:
koffsets += [l1, o1 + keystart]
voffsets += [l2, o2 + valuestart]
offsets = koffsets + voffsets
fileobj.write(struct.pack('Iiiiiii',
LE_MAGIC, # magic
0, # version
len(messages), # number of entries
7 * 4, # start of key index
7 * 4 + len(messages) * 8, # start of value index
0, 0 # size and offset of hash table
))
if PY3:
fileobj.write(array.array("i", offsets).tobytes())
else:
fileobj.write(array.array("i", offsets).tostring())
fileobj.write(ids + strs)
