# -*- coding: utf-8 -*-
from __future__ import with_statement
import re
import sys
import os
from contextlib import nested
reload(sys)
sys.setdefaultencoding('utf-8')
def clean(filename):
regex = re.compile(r'(?<!(&#\d{3};))(&#\d{3};){2}(?!(&#\d{3};))')
try:
os.rename(filename, "%s.old" %filename)
except:
print "Did not rename"
with nested(open(filename, "wb" ), open(filename+".old", "rb" )) as (destination, source):
counter = 0
for line in source:
rObj = re.search(regex, line)
counter += 1
print counter
if rObj is not None:
hexValues = [hex(int(rObj.group()[2:5])), hex(int(rObj.group()[8:11]))]
newChar = ''.join([chr(int(c, 16)) for c in hexValues]).decode('utf8')
newLine = re.sub(regex, newChar, line)
destination.write(newLine)
else:
destination.write(line)
os.remove("%s.old" %filename)
def usage():
print "Usage: python fix-xml.py relase, where release is for example 20091101"
sys.exit()
def main(argv):
if len(argv) == 0 or len(argv[0]) != 8:
usage()
try:
int(argv[0])
except ValueError:
usage()
sys.exit()
release = argv[0]
filename = 'discogs_%s_labels.xml' % release
clean(filename)
filename = 'discogs_%s_releases.xml' % release
clean(filename)
filename = 'discogs_%s_artists.xml' % release
clean(filename)
if __name__ == '__main__':
main(sys.argv[1:])