1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123
|
# -*- coding: utf-8 -*-
#
# Copyright (C) 2007 Edgewall Software
# All rights reserved.
#
# This software is licensed as described in the file COPYING, which
# you should have received as part of this distribution. The terms
# are also available at http://babel.edgewall.org/wiki/License.
#
# This software consists of voluntary contributions made by many
# individuals. For the exact contribution history, see the revision
# history and logs, available at http://babel.edgewall.org/log/.
"""Low-level locale data access.
:note: The `Locale` class, which uses this module under the hood, provides a
more convenient interface for accessing the locale data.
"""
import os
import pickle
try:
import threading
except ImportError:
import dummy_threading as threading
__all__ = ['exists', 'load']
__docformat__ = 'restructuredtext en'
_cache = {}
_cache_lock = threading.RLock()
_dirname = os.path.join(os.path.dirname(__file__), 'localedata')
def exists(name):
"""Check whether locale data is available for the given locale.
:param name: the locale identifier string
:return: `True` if the locale data exists, `False` otherwise
:rtype: `bool`
"""
if name in _cache:
return True
return os.path.exists(os.path.join(_dirname, '%s.dat' % name))
def list():
"""Return a list of all locale identifiers for which locale data is
available.
:return: a list of locale identifiers (strings)
:rtype: `list`
:since: version 0.8.1
"""
return [stem for stem, extension in [
os.path.splitext(filename) for filename in os.listdir(_dirname)
] if extension == '.dat' and stem != 'root']
def load(name):
"""Load the locale data for the given locale.
The locale data is a dictionary that contains much of the data defined by
the Common Locale Data Repository (CLDR). This data is stored as a
collection of pickle files inside the ``babel`` package.
>>> d = load('en_US')
>>> d['languages']['sv']
u'Swedish'
Note that the results are cached, and subsequent requests for the same
locale return the same dictionary:
>>> d1 = load('en_US')
>>> d2 = load('en_US')
>>> d1 is d2
True
:param name: the locale identifier string (or "root")
:return: the locale data
:rtype: `dict`
:raise `IOError`: if no locale data file is found for the given locale
identifer, or one of the locales it inherits from
"""
_cache_lock.acquire()
try:
data = _cache.get(name)
if not data:
# Load inherited data
if name == 'root':
data = {}
else:
parts = name.split('_')
if len(parts) == 1:
parent = 'root'
else:
parent = '_'.join(parts[:-1])
data = load(parent).copy()
filename = os.path.join(_dirname, '%s.dat' % name)
fileobj = open(filename, 'rb')
try:
if name != 'root':
merge(data, pickle.load(fileobj))
else:
data = pickle.load(fileobj)
_cache[name] = data
finally:
fileobj.close()
return data
finally:
_cache_lock.release()
def merge(dict1, dict2):
"""Merge the data from `dict2` into the `dict1` dictionary, making copies
of nested dictionaries.
:param dict1: the dictionary to merge into
:param dict2: the dictionary containing the data that should be merged
"""
for key, value in dict2.items():
if value is not None:
if type(value) is dict:
dict1[key] = dict1.get(key, {}).copy()
merge(dict1[key], value)
else:
dict1[key] = value
|