nojhan/weboob-devel
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity

[imdb] tests ok, unicode ok

Browse source
This commit is contained in:
Julien Veyssier 2013-03-06 14:24:01 +01:00
commit 1262744aa6
3 changed files with 11 additions and 11 deletions
Download patch file Download diff file Expand all files Collapse all files

8
modules/imdb/pages.py
View file

@ -110,19 +110,19 @@ class PersonPage(BasePage):
td_overview = self.parser.select(self.document.getroot(),'td#overview-top',1)
descs = self.parser.select(td_overview,'span[itemprop=description]')
if len(descs) > 0:
short_biography = descs[0].text
short_biography = u'%s'%descs[0].text
rname_block = self.parser.select(td_overview,'div.txt-block h4.inline')
if len(rname_block) > 0 and "born" in rname_block[0].text.lower():
links = self.parser.select(rname_block[0].getparent(),'a')
for a in links:
href = a.attrib.get('href','').strip()
if href == 'bio':
real_name = a.text.strip()
real_name = u'%s'%a.text.strip()
elif 'birth_place' in href:
birth_place = a.text.lower().strip()
birth_place = u'%s'%a.text.lower().strip()
names = self.parser.select(td_overview,'h1[itemprop=name]')
if len(names) > 0:
name = names[0].text.strip()
name = u'%s'%names[0].text.strip()
times = self.parser.select(td_overview,'time[itemprop=birthDate]')
if len(times) > 0:
time = times[0].attrib.get('datetime','').split('-')