from __future__ import unicode_literals
import re
-import json
from .common import InfoExtractor
from ..utils import (
def _real_extract(self, url):
mobj = re.match(self._VALID_URL, url)
- video_id = mobj.group('id')
lang = mobj.group('lang')
+ video_id = mobj.group('id')
ref_xml_url = url.replace('/videos/', '/do_delegate/videos/')
ref_xml_url = ref_xml_url.replace('.html', ',view,asPlayerXml.xml')
formats = [{
'forma_id': q.attrib['quality'],
- 'url': q.text,
+ # The playpath starts at 'mp4:', if we don't manually
+ # split the url, rtmpdump will incorrectly parse them
+ 'url': q.text.split('mp4:', 1)[0],
+ 'play_path': 'mp4:' + q.text.split('mp4:', 1)[1],
'ext': 'flv',
'quality': 2 if q.attrib['quality'] == 'hd' else 1,
} for q in config.findall('./urls/url')]
return self._extract_from_webpage(webpage, video_id, lang)
def _extract_from_webpage(self, webpage, video_id, lang):
- json_url = self._html_search_regex(r'arte_vp_url="(.*?)"', webpage, 'json url')
+ json_url = self._html_search_regex(
+ [r'arte_vp_url=["\'](.*?)["\']', r'data-url=["\']([^"]+)["\']'],
+ webpage, 'json vp url')
return self._extract_from_json_url(json_url, video_id, lang)
def _extract_from_json_url(self, json_url, video_id, lang):
regexes = [r'VO?%s' % l, r'VO?.-ST%s' % l]
return any(re.match(r, f['versionCode']) for r in regexes)
# Some formats may not be in the same language as the url
+ # TODO: Might want not to drop videos that does not match requested language
+ # but to process those formats with lower precedence
formats = filter(_match_lang, all_formats)
- formats = list(formats) # in python3 filter returns an iterator
+ formats = list(formats) # in python3 filter returns an iterator
if not formats:
# Some videos are only available in the 'Originalversion'
# they aren't tagged as being in French or German
- if all(f['versionCode'] == 'VO' for f in all_formats):
- formats = all_formats
- else:
- raise ExtractorError(u'The formats list is empty')
+ # Sometimes there are neither videos of requested lang code
+ # nor original version videos available
+ # For such cases we just take all_formats as is
+ formats = all_formats
+ if not formats:
+ raise ExtractorError('The formats list is empty')
if re.match(r'[A-Z]Q', formats[0]['quality']) is not None:
def sort_key(f):
return ['HQ', 'MQ', 'EQ', 'SQ'].index(f['quality'])
else:
def sort_key(f):
+ versionCode = f.get('versionCode')
+ if versionCode is None:
+ versionCode = ''
return (
# Sort first by quality
- int(f.get('height',-1)),
- int(f.get('bitrate',-1)),
+ int(f.get('height', -1)),
+ int(f.get('bitrate', -1)),
# The original version with subtitles has lower relevance
- re.match(r'VO-ST(F|A)', f.get('versionCode', '')) is None,
+ re.match(r'VO-ST(F|A)', versionCode) is None,
# The version with sourds/mal subtitles has also lower relevance
- re.match(r'VO?(F|A)-STM\1', f.get('versionCode', '')) is None,
+ re.match(r'VO?(F|A)-STM\1', versionCode) is None,
# Prefer http downloads over m3u8
0 if f['url'].endswith('m3u8') else 1,
)
# It also uses the arte_vp_url url from the webpage to extract the information
class ArteTVCreativeIE(ArteTVPlus7IE):
IE_NAME = 'arte.tv:creative'
- _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/magazine?/(?P<id>.+)'
+ _VALID_URL = r'https?://creative\.arte\.tv/(?P<lang>fr|de)/(?:magazine?/)?(?P<id>[^?#]+)'
- _TEST = {
+ _TESTS = [{
'url': 'http://creative.arte.tv/de/magazin/agentur-amateur-corporate-design',
'info_dict': {
- 'id': '050489-002',
+ 'id': '72176',
'ext': 'mp4',
- 'title': 'Agentur Amateur / Agence Amateur #2 : Corporate Design',
+ 'title': 'Folge 2 - Corporate Design',
+ 'upload_date': '20131004',
},
- }
+ }, {
+ 'url': 'http://creative.arte.tv/fr/Monty-Python-Reunion',
+ 'info_dict': {
+ 'id': '160676',
+ 'ext': 'mp4',
+ 'title': 'Monty Python live (mostly)',
+ 'description': 'Événement ! Quarante-cinq ans après leurs premiers succès, les légendaires Monty Python remontent sur scène.\n',
+ 'upload_date': '20140805',
+ }
+ }]
class ArteTVFutureIE(ArteTVPlus7IE):
_TEST = {
'url': 'http://future.arte.tv/fr/sujet/info-sciences#article-anchor-7081',
'info_dict': {
- 'id': '050940-003',
+ 'id': '5201',
'ext': 'mp4',
'title': 'Les champignons au secours de la planète',
+ 'upload_date': '20131101',
},
}