X-Git-Url: https://git.rapsys.eu/youtubedl/blobdiff_plain/7ceb2ec430c3363e0140a0519402428f36dc472e..eed74133841b43ce901ee0dd31c7dad5a234bd48:/youtube_dl/extractor/orf.py diff --git a/youtube_dl/extractor/orf.py b/youtube_dl/extractor/orf.py index cfca2a0..011e6be 100644 --- a/youtube_dl/extractor/orf.py +++ b/youtube_dl/extractor/orf.py @@ -1,54 +1,181 @@ # coding: utf-8 +from __future__ import unicode_literals -import re -import xml.etree.ElementTree import json +import re +import calendar +import datetime from .common import InfoExtractor from ..utils import ( - compat_urlparse, + HEADRequest, + unified_strdate, ExtractorError, - find_xpath_attr, ) -class ORFIE(InfoExtractor): - _VALID_URL = r'https?://tvthek.orf.at/(programs/.+?/episodes|topics/.+?)/(?P\d+)' + +class ORFTVthekIE(InfoExtractor): + IE_NAME = 'orf:tvthek' + IE_DESC = 'ORF TVthek' + _VALID_URL = r'https?://tvthek\.orf\.at/(?:programs/.+?/episodes|topics/.+?|program/[^/]+)/(?P\d+)' + + _TEST = { + 'url': 'http://tvthek.orf.at/program/matinee-Was-Sie-schon-immer-ueber-Klassik-wissen-wollten/7317210/Was-Sie-schon-immer-ueber-Klassik-wissen-wollten/7319746/Was-Sie-schon-immer-ueber-Klassik-wissen-wollten/7319747', + 'file': '7319747.mp4', + 'md5': 'bd803c5d8c32d3c64a0ea4b4eeddf375', + 'info_dict': { + 'title': 'Was Sie schon immer über Klassik wissen wollten', + 'description': 'md5:0ddf0d5f0060bd53f744edaa5c2e04a4', + 'duration': 3508, + 'upload_date': '20140105', + }, + 'skip': 'Blocked outside of Austria', + } def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) playlist_id = mobj.group('id') webpage = self._download_webpage(url, playlist_id) - flash_xml = self._search_regex('ORF.flashXML = \'(.+?)\'', webpage, u'flash xml') - flash_xml = compat_urlparse.parse_qs('xml='+flash_xml)['xml'][0] - flash_config = xml.etree.ElementTree.fromstring(flash_xml.encode('utf-8')) - playlist_json = self._search_regex(r'playlist\': \'(\[.*?\])\'', webpage, u'playlist').replace(r'\"','"') - playlist = json.loads(playlist_json) - - videos = [] - ns = '{http://tempuri.org/XMLSchema.xsd}' - xpath = '%(ns)sPlaylist/%(ns)sItems/%(ns)sItem' % {'ns': ns} - webpage_description = self._og_search_description(webpage) - for (i, (item, info)) in enumerate(zip(flash_config.findall(xpath), playlist), 1): - # Get best quality url - rtmp_url = None - for q in ['Q6A', 'Q4A', 'Q1A']: - video_url = find_xpath_attr(item, '%sVideoUrl' % ns, 'quality', q) - if video_url is not None: - rtmp_url = video_url.text - break - if rtmp_url is None: - raise ExtractorError(u'Couldn\'t get video url: %s' % info['id']) - description = self._html_search_regex( - r'id="playlist_entry_%s".*?

(.*?)

' % i, webpage, - u'description', default=webpage_description, flags=re.DOTALL) - videos.append({ + data_json = self._search_regex( + r'initializeAdworx\((.+?)\);\n', webpage, 'video info') + all_data = json.loads(data_json) + + def get_segments(all_data): + for data in all_data: + if data['name'] == 'Tracker::EPISODE_DETAIL_PAGE_OVER_PROGRAM': + return data['values']['segments'] + + sdata = get_segments(all_data) + if not sdata: + raise ExtractorError('Unable to extract segments') + + def quality_to_int(s): + m = re.search('([0-9]+)', s) + if m is None: + return -1 + return int(m.group(1)) + + entries = [] + for sd in sdata: + video_id = sd['id'] + formats = [{ + 'preference': -10 if fd['delivery'] == 'hls' else None, + 'format_id': '%s-%s-%s' % ( + fd['delivery'], fd['quality'], fd['quality_string']), + 'url': fd['src'], + 'protocol': fd['protocol'], + 'quality': quality_to_int(fd['quality']), + } for fd in sd['playlist_item_array']['sources']] + + # Check for geoblocking. + # There is a property is_geoprotection, but that's always false + geo_str = sd.get('geoprotection_string') + if geo_str: + try: + http_url = next( + f['url'] + for f in formats + if re.match(r'^https?://.*\.mp4$', f['url'])) + except StopIteration: + pass + else: + req = HEADRequest(http_url) + self._request_webpage( + req, video_id, + note='Testing for geoblocking', + errnote=(( + 'This video seems to be blocked outside of %s. ' + 'You may want to try the streaming-* formats.') + % geo_str), + fatal=False) + + self._sort_formats(formats) + + upload_date = unified_strdate(sd['created_date']) + entries.append({ '_type': 'video', - 'id': info['id'], - 'title': info['title'], - 'url': rtmp_url, - 'ext': 'flv', - 'description': description, - }) - - return videos + 'id': video_id, + 'title': sd['header'], + 'formats': formats, + 'description': sd.get('description'), + 'duration': int(sd['duration_in_seconds']), + 'upload_date': upload_date, + 'thumbnail': sd.get('image_full_url'), + }) + + return { + '_type': 'playlist', + 'entries': entries, + 'id': playlist_id, + } + + +# Audios on ORF radio are only available for 7 days, so we can't add tests. + + +class ORFOE1IE(InfoExtractor): + IE_NAME = 'orf:oe1' + IE_DESC = 'Radio Österreich 1' + _VALID_URL = r'http://oe1\.orf\.at/programm/(?P[0-9]+)' + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + show_id = mobj.group('id') + + data = self._download_json( + 'http://oe1.orf.at/programm/%s/konsole' % show_id, + show_id + ) + + timestamp = datetime.datetime.strptime('%s %s' % ( + data['item']['day_label'], + data['item']['time'] + ), '%d.%m.%Y %H:%M') + unix_timestamp = calendar.timegm(timestamp.utctimetuple()) + + return { + 'id': show_id, + 'title': data['item']['title'], + 'url': data['item']['url_stream'], + 'ext': 'mp3', + 'description': data['item'].get('info'), + 'timestamp': unix_timestamp + } + + +class ORFFM4IE(InfoExtractor): + IE_DESC = 'orf:fm4' + IE_DESC = 'radio FM4' + _VALID_URL = r'http://fm4\.orf\.at/7tage/?#(?P[0-9]+)/(?P\w+)' + + def _real_extract(self, url): + mobj = re.match(self._VALID_URL, url) + show_date = mobj.group('date') + show_id = mobj.group('show') + + data = self._download_json( + 'http://audioapi.orf.at/fm4/json/2.0/broadcasts/%s/4%s' % (show_date, show_id), + show_id + ) + + def extract_entry_dict(info, title, subtitle): + return { + 'id': info['loopStreamId'].replace('.mp3', ''), + 'url': 'http://loopstream01.apa.at/?channel=fm4&id=%s' % info['loopStreamId'], + 'title': title, + 'description': subtitle, + 'duration': (info['end'] - info['start']) / 1000, + 'timestamp': info['start'] / 1000, + 'ext': 'mp3' + } + + entries = [extract_entry_dict(t, data['title'], data['subtitle']) for t in data['streams']] + + return { + '_type': 'playlist', + 'id': show_id, + 'title': data['title'], + 'description': data['subtitle'], + 'entries': entries + } \ No newline at end of file