diff options
Diffstat (limited to 'youtube_dl/extractor/francetv.py')
| -rw-r--r-- | youtube_dl/extractor/francetv.py | 77 | 
1 files changed, 77 insertions, 0 deletions
diff --git a/youtube_dl/extractor/francetv.py b/youtube_dl/extractor/francetv.py new file mode 100644 index 000000000..f2b12c884 --- /dev/null +++ b/youtube_dl/extractor/francetv.py @@ -0,0 +1,77 @@ +# encoding: utf-8 +import re +import xml.etree.ElementTree + +from .common import InfoExtractor +from ..utils import ( +    compat_urlparse, +) + + +class FranceTVBaseInfoExtractor(InfoExtractor): +    def _extract_video(self, video_id): +        xml_desc = self._download_webpage( +            'http://www.francetvinfo.fr/appftv/webservices/video/' +            'getInfosOeuvre.php?id-diffusion=' +            + video_id, video_id, 'Downloading XML config') +        info = xml.etree.ElementTree.fromstring(xml_desc.encode('utf-8')) + +        manifest_url = info.find('videos/video/url').text +        video_url = manifest_url.replace('manifest.f4m', 'index_2_av.m3u8') +        video_url = video_url.replace('/z/', '/i/') +        thumbnail_path = info.find('image').text + +        return {'id': video_id, +                'ext': 'mp4', +                'url': video_url, +                'title': info.find('titre').text, +                'thumbnail': compat_urlparse.urljoin('http://pluzz.francetv.fr', thumbnail_path), +                'description': info.find('synopsis').text, +                } + + +class PluzzIE(FranceTVBaseInfoExtractor): +    IE_NAME = u'pluzz.francetv.fr' +    _VALID_URL = r'https?://pluzz\.francetv\.fr/videos/(.*?)\.html' + +    _TEST = { +        u'url': u'http://pluzz.francetv.fr/videos/allo_rufo_saison5_,88439064.html', +        u'file': u'88439064.mp4', +        u'info_dict': { +            u'title': u'AllĂ´ Rufo', +            u'description': u'md5:d909f1ebdf963814b65772aea250400e', +        }, +        u'params': { +            u'skip_download': True, +        }, +    } + +    def _real_extract(self, url): +        title = re.match(self._VALID_URL, url).group(1) +        webpage = self._download_webpage(url, title) +        video_id = self._search_regex( +            r'data-diffusion="(\d+)"', webpage, 'ID') +        return self._extract_video(video_id) + + +class FranceTvInfoIE(FranceTVBaseInfoExtractor): +    IE_NAME = u'francetvinfo.fr' +    _VALID_URL = r'https?://www\.francetvinfo\.fr/replay.*/(?P<title>.+).html' + +    _TEST = { +        u'url': u'http://www.francetvinfo.fr/replay-jt/france-3/soir-3/jt-grand-soir-3-lundi-26-aout-2013_393427.html', +        u'file': u'84981923.mp4', +        u'info_dict': { +            u'title': u'Soir 3', +        }, +        u'params': { +            u'skip_download': True, +        }, +    } + +    def _real_extract(self, url): +        mobj = re.match(self._VALID_URL, url) +        page_title = mobj.group('title') +        webpage = self._download_webpage(url, page_title) +        video_id = self._search_regex(r'id-video=(\d+?)"', webpage, u'video id') +        return self._extract_video(video_id)  | 
