#!/usr/bin/env python
# Need to ensure that the i18n framework is enabled
from django.conf import settings
settings.configure(USE_I18N = True)
from django.utils.translation import templatize
import re
import os
import sys
import getopt
from itertools import dropwhile
from tempfile import mkdtemp
from shutil import rmtree
pythonize_re = re.compile(r'\n\s*//')
msgfmt_re = re.compile(r'((?P<transl>\d+) translated messages?)?((, )?(?P<fuzzy>\d+) fuzzy translations?)?((, )?(?P<untransl>\d+) untranslated messages?)?\.')
def calculate_stats():
localedir = None
if os.path.isdir(os.path.join('conf', 'locale')):
#localedir = os.path.abspath(os.path.join('conf', 'locale'))
localedir = os.path.join('conf', 'locale')
elif os.path.isdir('locale'):
#localedir = os.path.abspath('locale')
localedir = 'locale'
else:
print "This script should be run from the django svn tree or your project or app tree."
print "If you did indeed run it from the svn checkout or your project or application,"
print "maybe you are just missing the conf/locale (in the django tree) or locale (for project"
print "and application) directory?."
sys.exit(1)
(opts, args) = getopt.getopt(sys.argv[1:], 'l:d:va')
lang = None
domain = 'django'
verbose = False
all = False
for o, v in opts:
if o == '-l':
lang = v
elif o == '-d':
domain = v
elif o == '-v':
verbose = True
elif o == '-a':
all = True
if domain not in ('django', 'djangojs'):
print "currently l10n-stats.py only supports domains 'django' and 'djangojs'"
sys.exit(1)
if (lang is None and not all) or domain is None:
print "usage: l10n-stats.py -l <language>"
print " or: l10n-stats.py -a"
sys.exit(1)
languages = []
if lang is not None:
languages.append(lang)
elif all:
languages = [el for el in os.listdir(localedir) if not el.startswith('.')]
if not languages:
sys.exit(0)
workdir = mkdtemp()
potfile = os.path.join(workdir, '%s.pot' % domain)
if os.path.exists(potfile):
os.unlink(potfile)
for (dirpath, dirnames, filenames) in os.walk("."):
for file in filenames:
if domain == 'djangojs' and file.endswith('.js'):
if verbose: sys.stdout.write('processing file %s in %s\n' % (file, dirpath))
data = open(os.path.join(dirpath, file), "rb").read()
data = pythonize_re.sub('\n#', data)
thefile = '%s.py' % file
open(os.path.join(dirpath, thefile), "wb").write(data)
cmd = 'xgettext %s -d %s -L Perl --keyword=gettext_noop --keyword=gettext_lazy --keyword=ngettext_lazy:1,2 --from-code UTF-8 -o - "%s"' % (
os.path.exists(potfile) and '--omit-header' or '', domain, os.path.join(dirpath, thefile))
(stdin, stdout, stderr) = os.popen3(cmd, 't')
msgs = stdout.read()
errors = stderr.read()
if errors:
sys.stderr.write('errors happened while running xgettext on %s\n' % file)
sys.stderr.write(errors)
rmtree(workdir, True)
sys.exit(8)
if msgs:
open(potfile, 'ab').write(msgs)
os.unlink(os.path.join(dirpath, thefile))
elif domain == 'django' and (file.endswith('.py') or file.endswith('.html')):
thefile = file
litfile = os.path.join(dirpath, file)
if file.endswith('.html'):
data = open(litfile, "rb").read()
thefile = '%s.py' % file
litdir = os.path.join(workdir, dirpath)
if not os.path.isdir(litdir):
os.makedirs(litdir)
litfile = os.path.join(litdir, thefile)
open(litfile, "wb").write(templatize(data))
if verbose:
sys.stdout.write('processing file %s in %s\n' % (file, dirpath))
cmd = 'xgettext -d %s -L Python --keyword=gettext_noop --keyword=gettext_lazy --keyword=ngettext_lazy:1,2 --keyword=ugettext_noop --keyword=ugettext_lazy --keyword=ungettext_lazy:1,2 --from-code UTF-8 -o - "%s"' % (
domain, litfile)
(stdin, stdout, stderr) = os.popen3(cmd, 't')
msgs = stdout.read()
errors = stderr.read()
if errors:
sys.stderr.write('errors happened while running xgettext on %s\n' % file)
sys.stderr.write(errors)
rmtree(workdir, True)
sys.exit(8)
if os.path.exists(potfile):
# Strip the header
msgs = '\n'.join(dropwhile(len, msgs.split('\n')))
else:
msgs = msgs.replace('charset=CHARSET', 'charset=UTF-8')
if msgs:
open(potfile, 'ab').write(msgs)
if os.path.exists(potfile):
(stdin, stdout, stderr) = os.popen3('msguniq --to-code=utf-8 "%s"' % potfile, 'b')
pot_msgs = stdout.read()
errors = stderr.read()
if errors:
sys.stderr.write('errors happened while running msguniqi\n')
sys.stderr.write(errors)
rmtree(workdir, True)
sys.exit(8)
open(potfile, 'w').write(pot_msgs)
else:
sys.exit(0)
for lang in languages:
basedir = os.path.join(localedir, lang, 'LC_MESSAGES')
if not os.path.isdir(basedir):
continue
dstdir = os.path.join(workdir, basedir)
if not os.path.isdir(dstdir):
os.makedirs(dstdir)
pofile = os.path.join(basedir, '%s.po' % domain)
dstfile = os.path.join(dstdir, '%s.po' % domain)
if os.path.exists(pofile):
(stdin, stdout, stderr) = os.popen3('msgmerge -q "%s" "%s"' % (pofile, potfile), 'b')
msgs = stdout.read()
errors = stderr.read()
if errors:
sys.stderr.write('errors happened while running msgmerge\n')
sys.stderr.write(errors)
rmtree(workdir, True)
sys.exit(8)
open(dstfile, 'wb').write(msgs)
else:
open(dstfile, 'wb').write(pot_msgs)
(stdin, stdout, stderr) = os.popen3('LC_ALL=C msgfmt --statistics -o - "%s"' % dstfile, 't')
dummy = stdout.read()
data = stderr.read()
mo = msgfmt_re.match(data)
if mo:
groups = mo.groupdict('0')
transl = int(groups['transl'])
fuzzy = int(groups['fuzzy'])
untransl = int(groups['untransl'])
total = transl + fuzzy + untransl
print("%s: translated: %d%%, fuzzy: %d%%, untranslated: %d%%" % (lang, transl*100/total, fuzzy*100/total, untransl*100/total))
#os.unlink(potfile)
#os.rmdir(workdir)
rmtree(workdir, True)
if __name__ == "__main__":
calculate_stats()