From: Sergey M․ Date: Sat, 13 Jan 2018 16:28:08 +0000 (+0700) Subject: [wdr] Rework extractors (closes #14598) X-Git-Url: http://git.bitcoin.ninja/?a=commitdiff_plain;h=54e8f62e01b54eeccd8313349f86ae541082704f;p=youtube-dl [wdr] Rework extractors (closes #14598) --- diff --git a/youtube_dl/extractor/extractors.py b/youtube_dl/extractor/extractors.py index 255df75fe..c82614bf9 100644 --- a/youtube_dl/extractor/extractors.py +++ b/youtube_dl/extractor/extractors.py @@ -991,7 +991,6 @@ from .stitcher import StitcherIE from .sport5 import Sport5IE from .sportbox import SportBoxEmbedIE from .sportdeutschland import SportDeutschlandIE -from .sportschau import SportschauIE from .sprout import SproutIE from .srgssr import ( SRGSSRIE, @@ -1289,6 +1288,7 @@ from .watchbox import WatchBoxIE from .watchindianporn import WatchIndianPornIE from .wdr import ( WDRIE, + WDRPageIE, WDRElefantIE, WDRMobileIE, ) diff --git a/youtube_dl/extractor/sportschau.py b/youtube_dl/extractor/sportschau.py deleted file mode 100644 index 0d7925a08..000000000 --- a/youtube_dl/extractor/sportschau.py +++ /dev/null @@ -1,38 +0,0 @@ -# coding: utf-8 -from __future__ import unicode_literals - -from .wdr import WDRBaseIE -from ..utils import get_element_by_attribute - - -class SportschauIE(WDRBaseIE): - IE_NAME = 'Sportschau' - _VALID_URL = r'https?://(?:www\.)?sportschau\.de/(?:[^/]+/)+video-?(?P[^/#?]+)\.html' - _TEST = { - 'url': 'http://www.sportschau.de/uefaeuro2016/videos/video-dfb-team-geht-gut-gelaunt-ins-spiel-gegen-polen-100.html', - 'info_dict': { - 'id': 'mdb-1140188', - 'display_id': 'dfb-team-geht-gut-gelaunt-ins-spiel-gegen-polen-100', - 'ext': 'mp4', - 'title': 'DFB-Team geht gut gelaunt ins Spiel gegen Polen', - 'description': 'Vor dem zweiten Gruppenspiel gegen Polen herrscht gute Stimmung im deutschen Team. Insbesondere Bastian Schweinsteiger strotzt vor Optimismus nach seinem Tor gegen die Ukraine.', - 'upload_date': '20160615', - }, - 'skip': 'Geo-restricted to Germany', - } - - def _real_extract(self, url): - video_id = self._match_id(url) - - webpage = self._download_webpage(url, video_id) - title = get_element_by_attribute('class', 'headline', webpage) - description = self._html_search_meta('description', webpage, 'description') - - info = self._extract_wdr_video(webpage, video_id) - - info.update({ - 'title': title, - 'description': description, - }) - - return info diff --git a/youtube_dl/extractor/wdr.py b/youtube_dl/extractor/wdr.py index 4871ae92b..6bf5aeaed 100644 --- a/youtube_dl/extractor/wdr.py +++ b/youtube_dl/extractor/wdr.py @@ -4,50 +4,52 @@ from __future__ import unicode_literals import re from .common import InfoExtractor +from ..compat import ( + compat_str, + compat_urlparse, +) from ..utils import ( determine_ext, ExtractorError, js_to_json, strip_jsonp, + try_get, unified_strdate, update_url_query, urlhandle_detect_ext, ) -class WDRBaseIE(InfoExtractor): - def _extract_jsonp_url(self, webpage, display_id): - # for wdr.de the data-extension is in a tag with the class "mediaLink" - # for wdr.de radio players, in a tag with the class "wdrrPlayerPlayBtn" - # for wdrmaus, in a tag with the class "videoButton" (previously a link - # to the page in a multiline "videoLink"-tag) - json_metadata = self._html_search_regex( - r'''(?sx)class= - (?: - (["\'])(?:mediaLink|wdrrPlayerPlayBtn|videoButton)\b.*?\1[^>]+| - (["\'])videoLink\b.*?\2[\s]*>\n[^\n]* - )data-extension=(["\'])(?P(?:(?!\3).)+)\3 - ''', - webpage, 'media link', default=None, group='data') - - if not json_metadata: - return +class WDRIE(InfoExtractor): + _VALID_URL = r'https?://deviceids-medp\.wdr\.de/ondemand/\d+/(?P\d+)\.js' + _TEST = { + 'url': 'http://deviceids-medp.wdr.de/ondemand/155/1557833.js', + 'info_dict': { + 'id': 'mdb-1140188', + 'display_id': 'dfb-team-geht-gut-gelaunt-ins-spiel-gegen-polen-100', + 'ext': 'mp4', + 'title': 'DFB-Team geht gut gelaunt ins Spiel gegen Polen', + 'description': 'Vor dem zweiten Gruppenspiel gegen Polen herrscht gute Stimmung im deutschen Team. Insbesondere Bastian Schweinsteiger strotzt vor Optimismus nach seinem Tor gegen die Ukraine.', + 'upload_date': '20160615', + }, + 'skip': 'Geo-restricted to Germany', + } - media_link_obj = self._parse_json(json_metadata, display_id, - transform_source=js_to_json) - return media_link_obj['mediaObj']['url'] + def _real_extract(self, url): + video_id = self._match_id(url) - def _extract_wdr_video(self, jsonp_url, display_id): metadata = self._download_json( - jsonp_url, display_id, transform_source=strip_jsonp) + url, video_id, transform_source=strip_jsonp) - metadata_tracker_data = metadata['trackerData'] - metadata_media_resource = metadata['mediaResource'] + is_live = metadata.get('mediaType') == 'live' + + tracker_data = metadata['trackerData'] + media_resource = metadata['mediaResource'] formats = [] # check if the metadata contains a direct URL to a file - for kind, media_resource in metadata_media_resource.items(): + for kind, media_resource in media_resource.items(): if kind not in ('dflt', 'alt'): continue @@ -58,13 +60,13 @@ class WDRBaseIE(InfoExtractor): ext = determine_ext(medium_url) if ext == 'm3u8': formats.extend(self._extract_m3u8_formats( - medium_url, display_id, 'mp4', 'm3u8_native', + medium_url, video_id, 'mp4', 'm3u8_native', m3u8_id='hls')) elif ext == 'f4m': manifest_url = update_url_query( medium_url, {'hdcore': '3.2.0', 'plugin': 'aasp-3.2.0.77.18'}) formats.extend(self._extract_f4m_formats( - manifest_url, display_id, f4m_id='hds', fatal=False)) + manifest_url, video_id, f4m_id='hds', fatal=False)) elif ext == 'smil': formats.extend(self._extract_smil_formats( medium_url, 'stream', fatal=False)) @@ -74,7 +76,7 @@ class WDRBaseIE(InfoExtractor): } if ext == 'unknown_video': urlh = self._request_webpage( - medium_url, display_id, note='Determining extension') + medium_url, video_id, note='Determining extension') ext = urlhandle_detect_ext(urlh) a_format['ext'] = ext formats.append(a_format) @@ -82,30 +84,30 @@ class WDRBaseIE(InfoExtractor): self._sort_formats(formats) subtitles = {} - caption_url = metadata_media_resource.get('captionURL') + caption_url = media_resource.get('captionURL') if caption_url: subtitles['de'] = [{ 'url': caption_url, 'ext': 'ttml', }] - title = metadata_tracker_data['trackerClipTitle'] + title = tracker_data['trackerClipTitle'] return { - 'id': metadata_tracker_data.get('trackerClipId', display_id), - 'display_id': display_id, - 'title': title, - 'alt_title': metadata_tracker_data.get('trackerClipSubcategory'), + 'id': tracker_data.get('trackerClipId', video_id), + 'title': self._live_title(title) if is_live else title, + 'alt_title': tracker_data.get('trackerClipSubcategory'), 'formats': formats, 'subtitles': subtitles, - 'upload_date': unified_strdate(metadata_tracker_data.get('trackerClipAirTime')), + 'upload_date': unified_strdate(tracker_data.get('trackerClipAirTime')), + 'is_live': is_live, } -class WDRIE(WDRBaseIE): +class WDRPageIE(InfoExtractor): _CURRENT_MAUS_URL = r'https?://(?:www\.)wdrmaus.de/(?:[^/]+/){1,2}[^/?#]+\.php5' - _PAGE_REGEX = r'/(?:mediathek/)?[^/]+/(?P[^/]+)/(?P.+)\.html' - _VALID_URL = r'(?Phttps?://(?:www\d\.)?wdr\d?\.de)' + _PAGE_REGEX + '|' + _CURRENT_MAUS_URL + _PAGE_REGEX = r'/(?:mediathek/)?(?:[^/]+/)*(?P[^/]+)\.html' + _VALID_URL = r'https?://(?:www\d?\.)?(?:wdr\d?|sportschau)\.de' + _PAGE_REGEX + '|' + _CURRENT_MAUS_URL _TESTS = [ { @@ -125,6 +127,7 @@ class WDRIE(WDRBaseIE): 'ext': 'ttml', }]}, }, + 'skip': 'HTTP Error 404: Not Found', }, { 'url': 'http://www1.wdr.de/mediathek/audio/wdr3/wdr3-gespraech-am-samstag/audio-schriftstellerin-juli-zeh-100.html', @@ -140,19 +143,17 @@ class WDRIE(WDRBaseIE): 'is_live': False, 'subtitles': {} }, + 'skip': 'HTTP Error 404: Not Found', }, { 'url': 'http://www1.wdr.de/mediathek/video/live/index.html', 'info_dict': { - 'id': 'mdb-103364', + 'id': 'mdb-1406149', 'ext': 'mp4', - 'display_id': 'index', - 'title': r're:^WDR Fernsehen im Livestream [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$', + 'title': r're:^WDR Fernsehen im Livestream \(nur in Deutschland erreichbar\) [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$', 'alt_title': 'WDR Fernsehen Live', - 'upload_date': None, - 'description': 'md5:ae2ff888510623bf8d4b115f95a9b7c9', + 'upload_date': '20150101', 'is_live': True, - 'subtitles': {} }, 'params': { 'skip_download': True, # m3u8 download @@ -160,19 +161,18 @@ class WDRIE(WDRBaseIE): }, { 'url': 'http://www1.wdr.de/mediathek/video/sendungen/aktuelle-stunde/aktuelle-stunde-120.html', - 'playlist_mincount': 8, + 'playlist_mincount': 7, 'info_dict': { - 'id': 'aktuelle-stunde/aktuelle-stunde-120', + 'id': 'aktuelle-stunde-120', }, }, { 'url': 'http://www.wdrmaus.de/aktuelle-sendung/index.php5', 'info_dict': { - 'id': 'mdb-1323501', + 'id': 'mdb-1552552', 'ext': 'mp4', 'upload_date': 're:^[0-9]{8}$', 'title': 're:^Die Sendung mit der Maus vom [0-9.]{10}$', - 'description': 'Die Seite mit der Maus -', }, 'skip': 'The id changes from week to week because of the new episode' }, @@ -184,7 +184,6 @@ class WDRIE(WDRBaseIE): 'ext': 'mp4', 'upload_date': '20130919', 'title': 'Sachgeschichte - Achterbahn ', - 'description': 'Die Seite mit der Maus -', }, }, { @@ -192,83 +191,100 @@ class WDRIE(WDRBaseIE): # Live stream, MD5 unstable 'info_dict': { 'id': 'mdb-869971', - 'ext': 'flv', - 'title': 'COSMO Livestream', - 'description': 'md5:2309992a6716c347891c045be50992e4', + 'ext': 'mp4', + 'title': r're:^COSMO Livestream [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$', 'upload_date': '20160101', }, + 'params': { + 'skip_download': True, # m3u8 download + } + }, + { + 'url': 'http://www.sportschau.de/handballem2018/handball-nationalmannschaft-em-stolperstein-vorrunde-100.html', + 'info_dict': { + 'id': 'mdb-1556012', + 'ext': 'mp4', + 'title': 'DHB-Vizepräsident Bob Hanning - "Die Weltspitze ist extrem breit"', + 'upload_date': '20180111', + }, + 'params': { + 'skip_download': True, + }, + }, + { + 'url': 'http://www.sportschau.de/handballem2018/audio-vorschau---die-handball-em-startet-mit-grossem-favoritenfeld-100.html', + 'only_matching': True, } ] def _real_extract(self, url): mobj = re.match(self._VALID_URL, url) - url_type = mobj.group('type') - page_url = mobj.group('page_url') display_id = mobj.group('display_id') webpage = self._download_webpage(url, display_id) - jsonp_url = self._extract_jsonp_url(webpage, display_id) - info_dict = self._extract_wdr_video(jsonp_url, display_id) - - if not info_dict: - entries = [ - self.url_result(page_url + href[0], 'WDR') - for href in re.findall( - r']+data-extension=' % self._PAGE_REGEX, - webpage) - ] - - if entries: # Playlist page - return self.playlist_result(entries, playlist_id=display_id) - - raise ExtractorError('No downloadable streams found', expected=True) - - is_live = url_type == 'live' + entries = [] - if is_live: - info_dict.update({ - 'title': self._live_title(info_dict['title']), - 'upload_date': None, - }) - elif 'upload_date' not in info_dict: - info_dict['upload_date'] = unified_strdate(self._html_search_meta('DC.Date', webpage, 'upload date')) + # Article with several videos - info_dict.update({ - 'description': self._html_search_meta('Description', webpage), - 'is_live': is_live, - }) + # for wdr.de the data-extension is in a tag with the class "mediaLink" + # for wdr.de radio players, in a tag with the class "wdrrPlayerPlayBtn" + # for wdrmaus, in a tag with the class "videoButton" (previously a link + # to the page in a multiline "videoLink"-tag) + for mobj in re.finditer( + r'''(?sx)class= + (?: + (["\'])(?:mediaLink|wdrrPlayerPlayBtn|videoButton)\b.*?\1[^>]+| + (["\'])videoLink\b.*?\2[\s]*>\n[^\n]* + )data-extension=(["\'])(?P(?:(?!\3).)+)\3 + ''', webpage): + media_link_obj = self._parse_json( + mobj.group('data'), display_id, transform_source=js_to_json, + fatal=False) + if not media_link_obj: + continue + jsonp_url = try_get( + media_link_obj, lambda x: x['mediaObj']['url'], compat_str) + if jsonp_url: + entries.append(self.url_result(jsonp_url, ie=WDRIE.ie_key())) - return info_dict + # Playlist (e.g. https://www1.wdr.de/mediathek/video/sendungen/aktuelle-stunde/aktuelle-stunde-120.html) + if not entries: + entries = [ + self.url_result( + compat_urlparse.urljoin(url, mobj.group('href')), + ie=WDRPageIE.ie_key()) + for mobj in re.finditer( + r']+\bhref=(["\'])(?P(?:(?!\1).)+)\1[^>]+\bdata-extension=', + webpage) if re.match(self._PAGE_REGEX, mobj.group('href')) + ] + return self.playlist_result(entries, playlist_id=display_id) -class WDRElefantIE(WDRBaseIE): - _VALID_URL = r'https?://(?:www\.)wdrmaus.de/elefantenseite/#(?P.+)' - IE_NAME = 'wdr:elefant' - _TESTS = [ - { - 'url': 'http://www.wdrmaus.de/elefantenseite/#folge_ostern_2015', - 'info_dict': { - 'title': 'Folge Oster-Spezial 2015', - 'id': 'mdb-1088195', - 'ext': 'mp4', - 'age_limit': None, - 'upload_date': '20150406' - }, - 'params': { - 'skip_download' : True, - }, +class WDRElefantIE(InfoExtractor): + _VALID_URL = r'https?://(?:www\.)wdrmaus\.de/elefantenseite/#(?P.+)' + _TEST = { + 'url': 'http://www.wdrmaus.de/elefantenseite/#folge_ostern_2015', + 'info_dict': { + 'title': 'Folge Oster-Spezial 2015', + 'id': 'mdb-1088195', + 'ext': 'mp4', + 'age_limit': None, + 'upload_date': '20150406' }, - ] + 'params': { + 'skip_download': True, + }, + } def _real_extract(self, url): - mobj = re.match(self._VALID_URL, url) - display_id = mobj.group('display_id') + display_id = self._match_id(url) # Table of Contents seems to always be at this address, so fetch it directly. # The website fetches configurationJS.php5, which links to tableOfContentsJS.php5. table_of_contents = self._download_json( - 'https://www.wdrmaus.de/elefantenseite/data/tableOfContentsJS.php5', display_id) + 'https://www.wdrmaus.de/elefantenseite/data/tableOfContentsJS.php5', + display_id) if display_id not in table_of_contents: raise ExtractorError( 'No entry in site\'s table of contents for this URL. ' @@ -276,15 +292,13 @@ class WDRElefantIE(WDRBaseIE): expected=True) xml_metadata_path = table_of_contents[display_id]['xmlPath'] xml_metadata = self._download_xml( - 'https://www.wdrmaus.de/elefantenseite/' + xml_metadata_path, display_id) + 'https://www.wdrmaus.de/elefantenseite/' + xml_metadata_path, + display_id) zmdb_url_element = xml_metadata.find('./movie/zmdb_url') if zmdb_url_element is None: raise ExtractorError( - 'The URL looks valid, but no video was found. Note that download only works ' - 'on pages showing a single video, not on video selection pages.', - expected=True) - info_dict = self._extract_wdr_video(zmdb_url_element.text, display_id) - return info_dict + '%s is not a video' % display_id, expected=True) + return self.url_result(zmdb_url_element.text, ie=WDRIE.ie_key()) class WDRMobileIE(InfoExtractor):