w/youtube-dl
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

ArteTVIE: extract the video with the correct language

Browse source 
Some urls from the French version of the page could download the German version.

Also instead of extracting the json url from the webpage, build it to skip the download
This commit is contained in:
Jaime Marquínez Ferrándiz 2013-07-02 17:34:40 +02:00
parent 24a267b562
commit 9826925a20
1 changed files with 15 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

20
youtube_dl/extractor/arte.py
View file

@ -16,7 +16,7 @@ class ArteTvIE(InfoExtractor):
www.arte.tv/guide, the extraction process is different for each one. www.arte.tv/guide, the extraction process is different for each one.
The videos expire in 7 days, so we can't add tests. The videos expire in 7 days, so we can't add tests.
""" """
_EMISSION_URL = r'(?:http://)?www\.arte.tv/guide/(?:fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?' _EMISSION_URL = r'(?:http://)?www\.arte.tv/guide/(?P<lang>fr|de)/(?:(?:sendungen|emissions)/)?(?P<id>.*?)/(?P<name>.*?)(\?.*)?'
_VIDEOS_URL = r'(?:http://)?videos.arte.tv/(?:fr|de)/.*-(?P<id>.*?).html' _VIDEOS_URL = r'(?:http://)?videos.arte.tv/(?:fr|de)/.*-(?P<id>.*?).html'
_LIVE_URL = r'index-[0-9]+\.html$' _LIVE_URL = r'index-[0-9]+\.html$'
@ -57,10 +57,11 @@ class ArteTvIE(InfoExtractor):
mobj = re.match(self._EMISSION_URL, url) mobj = re.match(self._EMISSION_URL, url)
if mobj is not None: if mobj is not None:
name = mobj.group('name') name = mobj.group('name')
lang = mobj.group('lang')
# This is not a real id, it can be for example AJT for the news # This is not a real id, it can be for example AJT for the news
# http://www.arte.tv/guide/fr/emissions/AJT/arte-journal # http://www.arte.tv/guide/fr/emissions/AJT/arte-journal
video_id = mobj.group('id') video_id = mobj.group('id')
return self._extract_emission(url, video_id) return self._extract_emission(url, video_id, lang)
mobj = re.match(self._VIDEOS_URL, url) mobj = re.match(self._VIDEOS_URL, url)
if mobj is not None: if mobj is not None:
@ -72,10 +73,9 @@ class ArteTvIE(InfoExtractor):
# self.extractLiveStream(url) # self.extractLiveStream(url)
# return # return
def _extract_emission(self, url, video_id): def _extract_emission(self, url, video_id, lang):
"""Extract from www.arte.tv/guide""" """Extract from www.arte.tv/guide"""
webpage = self._download_webpage(url, video_id) json_url = 'http://org-www.arte.tv/papi/tvguide/videos/stream/player/F/%s_PLUS7-F/ALL/ALL.json' % video_id
json_url = self._html_search_regex(r'arte_vp_url="(.*?)"', webpage, 'json url')
json_info = self._download_webpage(json_url, video_id, 'Downloading info json') json_info = self._download_webpage(json_url, video_id, 'Downloading info json')
self.report_extraction(video_id) self.report_extraction(video_id)
@ -91,6 +91,16 @@ class ArteTvIE(InfoExtractor):
} }
formats = player_info['VSR'].values() formats = player_info['VSR'].values()
def _match_lang(f):
# Return true if that format is in the language of the url
if lang == 'fr':
l = 'F'
elif lang == 'de':
l = 'A'
regexes = [r'VO?%s' % l, r'V%s-ST.' % l]
return any(re.match(r, f['versionCode']) for r in regexes)
# Some formats may not be in the same language as the url
formats = filter(_match_lang, formats)
# We order the formats by quality # We order the formats by quality
formats = sorted(formats, key=lambda f: int(f['height'])) formats = sorted(formats, key=lambda f: int(f['height']))
# Pick the best quality # Pick the best quality