X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/33cd347759d6d999325ebf3c69b7ed5692c343b2..16a91194b542b099c141633fd95e77db4ff075ed:/youtube_dl/extractor/dreisat.py?ds=sidebyside diff --git a/youtube_dl/extractor/dreisat.py b/youtube_dl/extractor/dreisat.py index 64b4658..8d31258 100644 --- a/youtube_dl/extractor/dreisat.py +++ b/youtube_dl/extractor/dreisat.py @@ -1,85 +1,193 @@ -# coding: utf-8 +from __future__ import unicode_literals import re -import xml.etree.ElementTree from .common import InfoExtractor from ..utils import ( - determine_ext, + int_or_none, unified_strdate, + xpath_text, + determine_ext, + float_or_none, + ExtractorError, ) class DreiSatIE(InfoExtractor): IE_NAME = '3sat' - _VALID_URL = r'(?:http://)?(?:www\.)?3sat.de/mediathek/index.php\?(?:(?:mode|display)=[^&]+&)*obj=(?P[0-9]+)$' - _TEST = { - u"url": u"http://www.3sat.de/mediathek/index.php?obj=36983", - u'file': u'36983.webm', - u'md5': u'57c97d0469d71cf874f6815aa2b7c944', - u'info_dict': { - u"title": u"Kaffeeland Schweiz", - u"description": u"Über 80 Kaffeeröstereien liefern in der Schweiz das Getränk, in das das Land so vernarrt ist: Mehr als 1000 Tassen trinkt ein Schweizer pro Jahr. SCHWEIZWEIT nimmt die Kaffeekultur unter die...", - u"uploader": u"3sat", - u"upload_date": u"20130622" - } - } + _GEO_COUNTRIES = ['DE'] + _VALID_URL = r'https?://(?:www\.)?3sat\.de/mediathek/(?:(?:index|mediathek)\.php)?\?(?:(?:mode|display)=[^&]+&)*obj=(?P[0-9]+)' + _TESTS = [ + { + 'url': 'http://www.3sat.de/mediathek/index.php?mode=play&obj=45918', + 'md5': 'be37228896d30a88f315b638900a026e', + 'info_dict': { + 'id': '45918', + 'ext': 'mp4', + 'title': 'Waidmannsheil', + 'description': 'md5:cce00ca1d70e21425e72c86a98a56817', + 'uploader': 'SCHWEIZWEIT', + 'uploader_id': '100000210', + 'upload_date': '20140913' + }, + 'params': { + 'skip_download': True, # m3u8 downloads + } + }, + { + 'url': 'http://www.3sat.de/mediathek/mediathek.php?mode=play&obj=51066', + 'only_matching': True, + }, + ] + def _parse_smil_formats(self, smil, smil_url, video_id, namespace=None, f4m_params=None, transform_rtmp_url=None): + param_groups = {} + for param_group in smil.findall(self._xpath_ns('./head/paramGroup', namespace)): + group_id = param_group.get(self._xpath_ns( + 'id', 'http://www.w3.org/XML/1998/namespace')) + params = {} + for param in param_group: + params[param.get('name')] = param.get('value') + param_groups[group_id] = params - def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - video_id = mobj.group('id') - details_url = 'http://www.3sat.de/mediathek/xmlservice/web/beitragsDetails?ak=web&id=%s' % video_id - details_xml = self._download_webpage(details_url, video_id, note=u'Downloading video details') - details_doc = xml.etree.ElementTree.fromstring(details_xml.encode('utf-8')) - - thumbnail_els = details_doc.findall('.//teaserimage') - thumbnails = [{ - 'width': te.attrib['key'].partition('x')[0], - 'height': te.attrib['key'].partition('x')[2], - 'url': te.text, - } for te in thumbnail_els] - - information_el = details_doc.find('.//information') - video_title = information_el.find('./title').text - video_description = information_el.find('./detail').text - - details_el = details_doc.find('.//details') - video_uploader = details_el.find('./channel').text - upload_date = unified_strdate(details_el.find('./airtime').text) - - format_els = details_doc.findall('.//formitaet') - formats = [{ - 'format_id': fe.attrib['basetype'], - 'width': int(fe.find('./width').text), - 'height': int(fe.find('./height').text), - 'url': fe.find('./url').text, - 'filesize': int(fe.find('./filesize').text), - 'video_bitrate': int(fe.find('./videoBitrate').text), - '3sat_qualityname': fe.find('./quality').text, - } for fe in format_els - if not fe.find('./url').text.startswith('http://www.metafilegenerator.de/')] - - def _sortkey(format): - qidx = ['low', 'med', 'high', 'veryhigh'].index(format['3sat_qualityname']) - prefer_http = 1 if 'rtmp' in format['url'] else 0 - return (qidx, prefer_http, format['video_bitrate']) - formats.sort(key=_sortkey) - - info = { - '_type': 'video', + formats = [] + for video in smil.findall(self._xpath_ns('.//video', namespace)): + src = video.get('src') + if not src: + continue + bitrate = int_or_none(self._search_regex(r'_(\d+)k', src, 'bitrate', None)) or float_or_none(video.get('system-bitrate') or video.get('systemBitrate'), 1000) + group_id = video.get('paramGroup') + param_group = param_groups[group_id] + for proto in param_group['protocols'].split(','): + formats.append({ + 'url': '%s://%s' % (proto, param_group['host']), + 'app': param_group['app'], + 'play_path': src, + 'ext': 'flv', + 'format_id': '%s-%d' % (proto, bitrate), + 'tbr': bitrate, + }) + self._sort_formats(formats) + return formats + + def extract_from_xml_url(self, video_id, xml_url): + doc = self._download_xml( + xml_url, video_id, + note='Downloading video info', + errnote='Failed to download video info') + + status_code = xpath_text(doc, './status/statuscode') + if status_code and status_code != 'ok': + if status_code == 'notVisibleAnymore': + message = 'Video %s is not available' % video_id + else: + message = '%s returned error: %s' % (self.IE_NAME, status_code) + raise ExtractorError(message, expected=True) + + title = xpath_text(doc, './/information/title', 'title', True) + + urls = [] + formats = [] + for fnode in doc.findall('.//formitaeten/formitaet'): + video_url = xpath_text(fnode, 'url') + if not video_url or video_url in urls: + continue + urls.append(video_url) + + is_available = 'http://www.metafilegenerator' not in video_url + geoloced = 'static_geoloced_online' in video_url + if not is_available or geoloced: + continue + + format_id = fnode.attrib['basetype'] + format_m = re.match(r'''(?x) + (?P[^_]+)_(?P[^_]+)_(?P[^_]+)_ + (?P[^_]+)_(?P[^_]+)_(?P[^_]+) + ''', format_id) + + ext = determine_ext(video_url, None) or format_m.group('container') + + if ext == 'meta': + continue + elif ext == 'smil': + formats.extend(self._extract_smil_formats( + video_url, video_id, fatal=False)) + elif ext == 'm3u8': + # the certificates are misconfigured (see + # https://github.com/rg3/youtube-dl/issues/8665) + if video_url.startswith('https://'): + continue + formats.extend(self._extract_m3u8_formats( + video_url, video_id, 'mp4', 'm3u8_native', + m3u8_id=format_id, fatal=False)) + elif ext == 'f4m': + formats.extend(self._extract_f4m_formats( + video_url, video_id, f4m_id=format_id, fatal=False)) + else: + quality = xpath_text(fnode, './quality') + if quality: + format_id += '-' + quality + + abr = int_or_none(xpath_text(fnode, './audioBitrate'), 1000) + vbr = int_or_none(xpath_text(fnode, './videoBitrate'), 1000) + + tbr = int_or_none(self._search_regex( + r'_(\d+)k', video_url, 'bitrate', None)) + if tbr and vbr and not abr: + abr = tbr - vbr + + formats.append({ + 'format_id': format_id, + 'url': video_url, + 'ext': ext, + 'acodec': format_m.group('acodec'), + 'vcodec': format_m.group('vcodec'), + 'abr': abr, + 'vbr': vbr, + 'tbr': tbr, + 'width': int_or_none(xpath_text(fnode, './width')), + 'height': int_or_none(xpath_text(fnode, './height')), + 'filesize': int_or_none(xpath_text(fnode, './filesize')), + 'protocol': format_m.group('proto').lower(), + }) + + geolocation = xpath_text(doc, './/details/geolocation') + if not formats and geolocation and geolocation != 'none': + self.raise_geo_restricted(countries=self._GEO_COUNTRIES) + + self._sort_formats(formats) + + thumbnails = [] + for node in doc.findall('.//teaserimages/teaserimage'): + thumbnail_url = node.text + if not thumbnail_url: + continue + thumbnail = { + 'url': thumbnail_url, + } + thumbnail_key = node.get('key') + if thumbnail_key: + m = re.match('^([0-9]+)x([0-9]+)$', thumbnail_key) + if m: + thumbnail['width'] = int(m.group(1)) + thumbnail['height'] = int(m.group(2)) + thumbnails.append(thumbnail) + + upload_date = unified_strdate(xpath_text(doc, './/details/airtime')) + + return { 'id': video_id, - 'title': video_title, - 'formats': formats, - 'description': video_description, + 'title': title, + 'description': xpath_text(doc, './/information/detail'), + 'duration': int_or_none(xpath_text(doc, './/details/lengthSec')), 'thumbnails': thumbnails, - 'thumbnail': thumbnails[-1]['url'], - 'uploader': video_uploader, + 'uploader': xpath_text(doc, './/details/originChannelTitle'), + 'uploader_id': xpath_text(doc, './/details/originChannelId'), 'upload_date': upload_date, + 'formats': formats, } - # TODO: Remove when #980 has been merged - info['url'] = formats[-1]['url'] - info['ext'] = determine_ext(formats[-1]['url']) - - return info \ No newline at end of file + def _real_extract(self, url): + video_id = self._match_id(url) + details_url = 'http://www.3sat.de/mediathek/xmlservice/web/beitragsDetails?id=%s' % video_id + return self.extract_from_xml_url(video_id, details_url)