devtools/fix_po_encoding
author Sylvain Thénault <sylvain.thenault@logilab.fr>
Thu, 20 May 2010 20:50:00 +0200
changeset 5557 1a534c596bff
parent 0 b97547f5f1fa
permissions -rwxr-xr-x
[entity] continue cleanup of Entity/AnyEntity namespace * prefix by cw_ or _cw all '2nd zone' attributes / methods, other most commonly used will be done later (some methods used by client are made private anyway when they are for internal purpose anyway) * kill set_eid, use direct affectation instead * kill the auto-transmutation feature (pre_add_hook), hence releasing 3.9 will require releasing file cube as well

#!/usr/bin/python

"""usage: fix-po-encodings [filename...]
change the encoding of the po files passed as arguments to utf-8
"""
import sys
import re
import codecs

def change_encoding(filename, target='UTF-8'):
    fdesc = open(filename)
    data = fdesc.read()
    fdesc.close()
    encoding = find_encoding(data)
    if encoding == target:
        return
    data = fix_encoding(data, target)
    data = unicode(data, encoding)
    fdesc = codecs.open(filename, 'wb', encoding=target)
    fdesc.write(data)
    fdesc.close()

def find_encoding(data):
    regexp = re.compile(r'"Content-Type:.* charset=([a-zA-Z0-9-]+)\\n"', re.M)
    mo = regexp.search(data)
    if mo is None:
        raise ValueError('No encoding declaration')
    return mo.group(1)

def fix_encoding(data, target_encoding):
    regexp = re.compile(r'("Content-Type:.* charset=)(.*)(\\n")', re.M)
    return regexp.sub(r'\1%s\3' % target_encoding, data)
    


for filename in sys.argv[1:]:
    print filename
    change_encoding(filename)