0x2620/python-ox
2
1
Fork 0
Code Issues Pull requests Releases Wiki Activity

fix wikipedia movie parser

Browse source
This commit is contained in:
j 2013-12-22 13:38:43 +05:30
parent 5c1ab13749
commit 2abe99c89f
1 changed files with 14 additions and 14 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

28
ox/web/wikipedia.py
View file

@ -58,21 +58,21 @@ def get_movie_data(wikipedia_url):
key = _key key = _key
if key[0] == '|': if key[0] == '|':
key = key[1:] key = key[1:]
value = d[1].strip() key = key.strip()
value = value.replace('<!-- see WP:ALT -->', '') value = d[1].strip()
if '<br>' in value: value = value.replace('<!-- see WP:ALT -->', '')
value = value.split('<br>') if '<br>' in value:
key = key.strip() value = value.split('<br>')
if value: if value:
if key in filmbox: if key in filmbox:
if isinstance(value, list) and isinstance(filmbox[key], basestring): if isinstance(value, list) and isinstance(filmbox[key], basestring):
filmbox[key] = [filmbox[key]] + value filmbox[key] = [filmbox[key]] + value
else:
filmbox[key] += value
if isinstance(filmbox[key], list):
filmbox[key] = [k for k in filmbox[key] if k]
else: else:
filmbox[key] += value filmbox[key] = value
if isinstance(filmbox[key], list):
filmbox[key] = [k for k in filmbox[key] if k]
else:
filmbox[key] = value
if not filmbox_data: if not filmbox_data:
return filmbox return filmbox
if 'amg_id' in filmbox and not filmbox['amg_id'].isdigit(): if 'amg_id' in filmbox and not filmbox['amg_id'].isdigit():