"""Internationalization and localization support.
This module provides internationalization (I18N) and localization (L10N)
support for your Python programs by providing an interface to the GNU gettext
message catalog library.
I18N refers to the operation by which a program is made aware of multiple
languages. L10N refers to the adaptation of your program, once
internationalized, to the local language and cultural habits.
"""
# This module represents the integration of work, contributions, feedback, and
# suggestions from the following people:
#
# Martin von Loewis, who wrote the initial implementation of the underlying
# C-based libintlmodule (later renamed _gettext), along with a skeletal
# gettext.py implementation.
#
# Peter Funk, who wrote fintl.py, a fairly complete wrapper around intlmodule,
# which also included a pure-Python implementation to read .mo files if
# intlmodule wasn't available.
#
# James Henstridge, who also wrote a gettext.py module, which has some
# interesting, but currently unsupported experimental features: the notion of
# a Catalog class and instances, and the ability to add to a catalog file via
# a Python API.
#
# Barry Warsaw integrated these modules, wrote the .install() API and code,
# and conformed all C and Python code to Python's coding standards.
#
# Francois Pinard and Marc-Andre Lemburg also contributed valuably to this
# module.
#
# J. David Ibanez implemented plural forms. Bruno Haible fixed some bugs.
#
# TODO:
# - Lazy loading of .mo files. Currently the entire catalog is loaded into
# memory, but that's probably bad for large translated programs. Instead,
# the lexical sort of original strings in GNU .mo files should be exploited
# to do binary searches and lazy initializations. Or you might want to use
# the undocumented double-hash algorithm for .mo files with hash tables, but
# you'll need to study the GNU gettext code to do this.
#
# - Support Solaris .mo file formats. Unfortunately, we've been unable to
# find this format documented anywhere.
import locale, copy, io, os, re, struct, sys
from errno import ENOENT
__all__ = ['NullTranslations', 'GNUTranslations', 'Catalog',
'find', 'translation', 'install', 'textdomain', 'bindtextdomain',
'bind_textdomain_codeset',
'dgettext', 'dngettext', 'gettext', 'lgettext', 'ldgettext',
'ldngettext', 'lngettext', 'ngettext',
]
_default_localedir = os.path.join(sys.base_prefix, 'share', 'locale')
# Expression parsing for plural form selection.
#
# The gettext library supports a small subset of C syntax. The only
# incompatible difference is that integer literals starting with zero are
# decimal.
#
# https://www.gnu.org/software/gettext/manual/gettext.html#Plural-forms
# http://git.savannah.gnu.org/cgit/gettext.git/tree/gettext-runtime/intl/plural.y
_token_pattern = re.compile(r"""
(?P[ \t]+) | # spaces and horizontal tabs
(?P[0-9]+\b) | # decimal integer
(?Pn\b) | # only n is allowed
(?P[()]) |
(?P[-*/%+?:]|[>,
# =, ==, !=, &&, ||,
# ? :
# unary and bitwise ops
# not allowed
(?P\w+|.) # invalid token
""", re.VERBOSE|re.DOTALL)
def _tokenize(plural):
for mo in re.finditer(_token_pattern, plural):
kind = mo.lastgroup
if kind == 'WHITESPACES':
continue
value = mo.group(kind)
if kind == 'INVALID':
raise ValueError('invalid token in plural form: %s' % value)
yield value
yield ''
def _error(value):
if value:
return ValueError('unexpected token in plural form: %s' % value)
else:
return ValueError('unexpected end of plural form')
_binary_ops = (
('||',),
('&&',),
('==', '!='),
('', '='),
('+', '-'),
('*', '/', '%'),
)
_binary_ops = {op: i for i, ops in enumerate(_binary_ops, 1) for op in ops}
_c2py_ops = {'||': 'or', '&&': 'and', '/': '//'}
def _parse(tokens, priority=-1):
result = ''
nexttok = next(tokens)
while nexttok == '!':
result += 'not '
nexttok = next(tokens)
if nexttok == '(':
sub, nexttok = _parse(tokens)
result = '%s(%s)' % (result, sub)
if nexttok != ')':
raise ValueError('unbalanced parenthesis in plural form')
elif nexttok == 'n':
result = '%s%s' % (result, nexttok)
else:
try:
value = int(nexttok, 10)
except ValueError:
raise _error(nexttok) from None
result = '%s%d' % (result, value)
nexttok = next(tokens)
j = 100
while nexttok in _binary_ops:
i = _binary_ops[nexttok]
if i < priority:
break
# Break chained comparisons
if i in (3, 4) and j in (3, 4): # '==', '!=', '', '='
result = '(%s)' % result
# Replace some C operators by their Python equivalents
op = _c2py_ops.get(nexttok, nexttok)
right, nexttok = _parse(tokens, i + 1)
result = '%s %s %s' % (result, op, right)
j = i
if j == priority == 4: # '', '='
result = '(%s)' % result
if nexttok == '?' and priority 1000:
raise ValueError('plural form expression is too long')
try:
result, nexttok = _parse(_tokenize(plural))
if nexttok:
raise _error(nexttok)
depth = 0
for c in result:
if c == '(':
depth += 1
if depth > 20:
# Python compiler limit is about 90.
# The most complex example has 2.
raise ValueError('plural form expression is too complex')
elif c == ')':
depth -= 1
ns = {'_as_int': _as_int}
exec('''if True:
def func(n):
if not isinstance(n, int):
n = _as_int(n)
return int(%s)
''' % result, ns)
return ns['func']
except RecursionError:
# Recursion error can be raised in _parse() or exec().
raise ValueError('plural form expression is too complex')
def _expand_lang(loc):
loc = locale.normalize(loc)
COMPONENT_CODESET = 1 = 0:
territory = loc[pos:]
loc = loc[:pos]
mask |= COMPONENT_TERRITORY
else:
territory = ''
language = loc
ret = []
for i in range(mask+1):
if not (i & ~mask): # if all components for this combo exist ...
val = language
if i & COMPONENT_TERRITORY: val += territory
if i & COMPONENT_CODESET: val += codeset
if i & COMPONENT_MODIFIER: val += modifier
ret.append(val)
ret.reverse()
return ret
class NullTranslations:
def __init__(self, fp=None):
self._info = {}
self._charset = None
self._output_charset = None
self._fallback = None
if fp is not None:
self._parse(fp)
def _parse(self, fp):
pass
def add_fallback(self, fallback):
if self._fallback:
self._fallback.add_fallback(fallback)
else:
self._fallback = fallback
def gettext(self, message):
if self._fallback:
return self._fallback.gettext(message)
return message
def lgettext(self, message):
if self._fallback:
return self._fallback.lgettext(message)
return message
def ngettext(self, msgid1, msgid2, n):
if self._fallback:
return self._fallback.ngettext(msgid1, msgid2, n)
if n == 1:
return msgid1
else:
return msgid2
def lngettext(self, msgid1, msgid2, n):
if self._fallback:
return self._fallback.lngettext(msgid1, msgid2, n)
if n == 1:
return msgid1
else:
return msgid2
def info(self):
return self._info
def charset(self):
return self._charset
def output_charset(self):
return self._output_charset
def set_output_charset(self, charset):
self._output_charset = charset
def install(self, names=None):
import builtins
builtins.__dict__['_'] = self.gettext
if hasattr(names, "__contains__"):
if "gettext" in names:
builtins.__dict__['gettext'] = builtins.__dict__['_']
if "ngettext" in names:
builtins.__dict__['ngettext'] = self.ngettext
if "lgettext" in names:
builtins.__dict__['lgettext'] = self.lgettext
if "lngettext" in names:
builtins.__dict__['lngettext'] = self.lngettext
class GNUTranslations(NullTranslations):
# Magic number of .mo files
LE_MAGIC = 0x950412de
BE_MAGIC = 0xde120495
# Acceptable .mo versions
VERSIONS = (0, 1)
def _get_versions(self, version):
"""Returns a tuple of major version, minor version"""
return (version >> 16, version & 0xffff)
def _parse(self, fp):
"""Override this method to support alternative .mo formats."""
unpack = struct.unpack
filename = getattr(fp, 'name', '')
# Parse the .mo file header, which consists of 5 little endian 32
# bit words.
self._catalog = catalog = {}
self.plural = lambda n: int(n != 1) # germanic plural by default
buf = fp.read()
buflen = len(buf)
# Are we big endian or little endian?
magic = unpack('