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

fall back to storyline for summary

Browse source
This commit is contained in:
j 2019-11-15 14:51:32 +01:00
parent cef85fc4de
commit 03c1191550
1 changed files with 9 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

10
ox/web/imdb.py
View file

@ -199,6 +199,11 @@ class Imdb(SiteParser):
'summary': zebra_table('Plot Summary', more=[ 'summary': zebra_table('Plot Summary', more=[
'<p>(.*?)<em' '<p>(.*?)<em'
]), ]),
'storyline': {
'page': '',
're': '<h2>Storyline</h2>.*?<p>(.*?)</p>',
'type': 'string'
},
'posterId': { 'posterId': {
'page': 'reference', 'page': 'reference',
're': '<img.*?class="titlereference-primary-image".*?src="(.*?)".*?>', 're': '<img.*?class="titlereference-primary-image".*?src="(.*?)".*?>',
@ -517,10 +522,13 @@ class Imdb(SiteParser):
]) ])
if self['releasedate'] == 'x': if self['releasedate'] == 'x':
del self['releasedate'] del self['releasedate']
if 'summary' not in self and 'storyline' in self:
self['summary'] = self.pop('storyline')
if 'summary' in self: if 'summary' in self:
if isinstance(self['summary'], list): if isinstance(self['summary'], list):
self['summary'] = self['summary'][0] self['summary'] = self['summary'][0]
self['summary'] = self['summary'].split('</p')[0].strip() self['summary'] = strip_tags(self['summary'].split('</p')[0]).split(' Written by\n')[0].strip()
if 'credits' in self: if 'credits' in self:
credits = [ credits = [