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

imdb parser fixes

Browse source
This commit is contained in:
j 2010-07-08 10:03:57 +02:00
parent 2f5c0ad31b
commit 87e73406eb
1 changed files with 6 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
ox/web/imdb.py
View file

@ -137,7 +137,7 @@ class Imdb(SiteParser):
}, },
'year': { 'year': {
'page': 'combined', 'page': 'combined',
're': '<a href="/year/(\d{4})/">', 're': '<meta name="og:title" content=".*?\((\d{4})\)"',
'type': 'int' 'type': 'int'
} }
} }
@ -146,13 +146,16 @@ class Imdb(SiteParser):
self.baseUrl = "http://www.imdb.com/title/tt%s/" % id self.baseUrl = "http://www.imdb.com/title/tt%s/" % id
super(Imdb, self).__init__() super(Imdb, self).__init__()
if 'runtime' in self: if 'runtime' in self and self['runtime']:
if 'min' in self['runtime']: base=60 if 'min' in self['runtime']: base=60
else: base=1 else: base=1
self['runtime'] = int(findRe(self['runtime'], '([0-9]+)')) * base self['runtime'] = int(findRe(self['runtime'], '([0-9]+)')) * base
else:
self['runtime'] = 0
if 'connections' in self: if 'connections' in self:
cc={} cc={}
if len(self['connections']) == 2 and isinstance(self['connections'][0], basestring):
self['connections'] = [self['connections']]
for rel, data in self['connections']: for rel, data in self['connections']:
cc[unicode(rel)] = re.compile('<a href="/title/tt(\d{7})/">').findall(data) cc[unicode(rel)] = re.compile('<a href="/title/tt(\d{7})/">').findall(data)
self['connections'] = cc self['connections'] = cc