logo

youtube-dl

[mirror] Download/Watch videos from video hostersgit clone https://hacktivis.me/git/mirror/youtube-dl.git

ign.py (16256B)

    

  1. # coding: utf-8
  3. from __future__ import unicode_literals
  5. import re
  7. from .common import InfoExtractor
  8. from ..compat import (
  9.     compat_filter as filter,
  10.     compat_HTTPError,
  11.     compat_parse_qs,
  12.     compat_urlparse,
  13. )
  14. from ..utils import (
  15.     determine_ext,
  16.     error_to_compat_str,
  17.     extract_attributes,
  18.     ExtractorError,
  19.     int_or_none,
  20.     merge_dicts,
  21.     orderedSet,
  22.     parse_iso8601,
  23.     strip_or_none,
  24.     traverse_obj,
  25.     url_or_none,
  26.     urljoin,
  27. )
  30. class IGNBaseIE(InfoExtractor):
  31.     def _call_api(self, slug):
  32.         return self._download_json(
  33.             'http://apis.ign.com/{0}/v3/{0}s/slug/{1}'.format(self._PAGE_TYPE, slug), slug)
  35.     def _checked_call_api(self, slug):
  36.         try:
  37.             return self._call_api(slug)
  38.         except ExtractorError as e:
  39.             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 404:
  40.                 e.cause.args = e.cause.args or [
  41.                     e.cause.geturl(), e.cause.getcode(), e.cause.reason]
  42.                 raise ExtractorError(
  43.                     'Content not found: expired?', cause=e.cause,
  44.                     expected=True)
  45.             raise
  47.     def _extract_video_info(self, video, fatal=True):
  48.         video_id = video['videoId']
  50.         formats = []
  51.         refs = traverse_obj(video, 'refs', expected_type=dict) or {}
  53.         m3u8_url = url_or_none(refs.get('m3uUrl'))
  54.         if m3u8_url:
  55.             formats.extend(self._extract_m3u8_formats(
  56.                 m3u8_url, video_id, 'mp4', 'm3u8_native',
  57.                 m3u8_id='hls', fatal=False))
  59.         f4m_url = url_or_none(refs.get('f4mUrl'))
  60.         if f4m_url:
  61.             formats.extend(self._extract_f4m_formats(
  62.                 f4m_url, video_id, f4m_id='hds', fatal=False))
  64.         for asset in (video.get('assets') or []):
  65.             asset_url = url_or_none(asset.get('url'))
  66.             if not asset_url:
  67.                 continue
  68.             formats.append({
  69.                 'url': asset_url,
  70.                 'tbr': int_or_none(asset.get('bitrate'), 1000),
  71.                 'fps': int_or_none(asset.get('frame_rate')),
  72.                 'height': int_or_none(asset.get('height')),
  73.                 'width': int_or_none(asset.get('width')),
  74.             })
  76.         mezzanine_url = traverse_obj(
  77.             video, ('system', 'mezzanineUrl'), expected_type=url_or_none)
  78.         if mezzanine_url:
  79.             formats.append({
  80.                 'ext': determine_ext(mezzanine_url, 'mp4'),
  81.                 'format_id': 'mezzanine',
  82.                 'preference': 1,
  83.                 'url': mezzanine_url,
  84.             })
  86.         if formats or fatal:
  87.             self._sort_formats(formats)
  88.         else:
  89.             return
  91.         thumbnails = traverse_obj(
  92.             video, ('thumbnails', Ellipsis, {'url': 'url'}), expected_type=url_or_none)
  93.         tags = traverse_obj(
  94.             video, ('tags', Ellipsis, 'displayName'),
  95.             expected_type=lambda x: x.strip() or None)
  97.         metadata = traverse_obj(video, 'metadata', expected_type=dict) or {}
  98.         title = traverse_obj(
  99.             metadata, 'longTitle', 'title', 'name',
  100.             expected_type=lambda x: x.strip() or None)
  102.         return {
  103.             'id': video_id,
  104.             'title': title,
  105.             'description': strip_or_none(metadata.get('description')),
  106.             'timestamp': parse_iso8601(metadata.get('publishDate')),
  107.             'duration': int_or_none(metadata.get('duration')),
  108.             'thumbnails': thumbnails,
  109.             'formats': formats,
  110.             'tags': tags,
  111.         }
  113.     # yt-dlp shim
  114.     @classmethod
  115.     def _extract_from_webpage(cls, url, webpage):
  116.         for embed_url in orderedSet(
  117.                 cls._extract_embed_urls(url, webpage) or [], lazy=True):
  118.             yield cls.url_result(embed_url, None if cls._VALID_URL is False else cls)
  121. class IGNIE(IGNBaseIE):
  122.     """
  123.     Extractor for some of the IGN sites, like www.ign.com, es.ign.com de.ign.com.
  124.     Some videos of it.ign.com are also supported
  125.     """
  126.     _VIDEO_PATH_RE = r'/(?:\d{4}/\d{2}/\d{2}/)?(?P<id>.+?)'
  127.     _PLAYLIST_PATH_RE = r'(?:/?\?(?P<filt>[^&#]+))?'
  128.     _VALID_URL = (
  129.         r'https?://(?:.+?\.ign|www\.pcmag)\.com/videos(?:%s)'
  130.         % '|'.join((_VIDEO_PATH_RE + r'(?:[/?&#]|$)', _PLAYLIST_PATH_RE)))
  131.     IE_NAME = 'ign.com'
  132.     _PAGE_TYPE = 'video'
  134.     _TESTS = [{
  135.         'url': 'http://www.ign.com/videos/2013/06/05/the-last-of-us-review',
  136.         'md5': 'd2e1586d9987d40fad7867bf96a018ea',
  137.         'info_dict': {
  138.             'id': '8f862beef863986b2785559b9e1aa599',
  139.             'ext': 'mp4',
  140.             'title': 'The Last of Us Review',
  141.             'description': 'md5:c8946d4260a4d43a00d5ae8ed998870c',
  142.             'timestamp': 1370440800,
  143.             'upload_date': '20130605',
  144.             'tags': 'count:9',
  145.         },
  146.         'params': {
  147.             'nocheckcertificate': True,
  148.         },
  149.     }, {
  150.         'url': 'http://www.pcmag.com/videos/2015/01/06/010615-whats-new-now-is-gogo-snooping-on-your-data',
  151.         'md5': 'f1581a6fe8c5121be5b807684aeac3f6',
  152.         'info_dict': {
  153.             'id': 'ee10d774b508c9b8ec07e763b9125b91',
  154.             'ext': 'mp4',
  155.             'title': 'What\'s New Now: Is GoGo Snooping on Your Data?',
  156.             'description': 'md5:817a20299de610bd56f13175386da6fa',
  157.             'timestamp': 1420571160,
  158.             'upload_date': '20150106',
  159.             'tags': 'count:4',
  160.         },
  161.         'skip': '404 Not Found',
  162.     }, {
  163.         'url': 'https://www.ign.com/videos/is-a-resident-evil-4-remake-on-the-way-ign-daily-fix',
  164.         'only_matching': True,
  165.     }]
  167.     @classmethod
  168.     def _extract_embed_urls(cls, url, webpage):
  169.         grids = re.findall(
  170.             r'''(?s)<section\b[^>]+\bclass\s*=\s*['"](?:[\w-]+\s+)*?content-feed-grid(?!\B|-)[^>]+>(.+?)</section[^>]*>''',
  171.             webpage)
  172.         return filter(None,
  173.                       (urljoin(url, m.group('path')) for m in re.finditer(
  174.                           r'''<a\b[^>]+\bhref\s*=\s*('|")(?P<path>/videos%s)\1'''
  175.                           % cls._VIDEO_PATH_RE, grids[0] if grids else '')))
  177.     def _real_extract(self, url):
  178.         m = re.match(self._VALID_URL, url)
  179.         display_id = m.group('id')
  180.         if display_id:
  181.             return self._extract_video(url, display_id)
  182.         display_id = m.group('filt') or 'all'
  183.         return self._extract_playlist(url, display_id)
  185.     def _extract_playlist(self, url, display_id):
  186.         webpage = self._download_webpage(url, display_id)
  188.         return self.playlist_result(
  189.             (self.url_result(u, ie=self.ie_key())
  190.              for u in self._extract_embed_urls(url, webpage)),
  191.             playlist_id=display_id)
  193.     def _extract_video(self, url, display_id):
  194.         display_id = self._match_id(url)
  195.         video = self._checked_call_api(display_id)
  197.         info = self._extract_video_info(video)
  199.         return merge_dicts({
  200.             'display_id': display_id,
  201.         }, info)
  204. class IGNVideoIE(IGNBaseIE):
  205.     _VALID_URL = r'https?://.+?\.ign\.com/(?:[a-z]{2}/)?[^/]+/(?P<id>\d+)/(?:video|trailer)/'
  206.     _TESTS = [{
  207.         'url': 'http://me.ign.com/en/videos/112203/video/how-hitman-aims-to-be-different-than-every-other-s',
  208.         'md5': 'dd9aca7ed2657c4e118d8b261e5e9de1',
  209.         'info_dict': {
  210.             'id': 'e9be7ea899a9bbfc0674accc22a36cc8',
  211.             'ext': 'mp4',
  212.             'title': 'How Hitman Aims to Be Different Than Every Other Stealth Game - NYCC 2015',
  213.             'description': 'Taking out assassination targets in Hitman has never been more stylish.',
  214.             'timestamp': 1444665600,
  215.             'upload_date': '20151012',
  216.         },
  217.         'expected_warnings': ['HTTP Error 400: Bad Request'],
  218.     }, {
  219.         'url': 'http://me.ign.com/ar/angry-birds-2/106533/video/lrd-ldyy-lwl-lfylm-angry-birds',
  220.         'only_matching': True,
  221.     }, {
  222.         # Youtube embed
  223.         'url': 'https://me.ign.com/ar/ratchet-clank-rift-apart/144327/trailer/embed',
  224.         'only_matching': True,
  225.     }, {
  226.         # Twitter embed
  227.         'url': 'http://adria.ign.com/sherlock-season-4/9687/trailer/embed',
  228.         'only_matching': True,
  229.     }, {
  230.         # Vimeo embed
  231.         'url': 'https://kr.ign.com/bic-2018/3307/trailer/embed',
  232.         'only_matching': True,
  233.     }]
  235.     def _real_extract(self, url):
  236.         video_id = self._match_id(url)
  237.         parsed_url = compat_urlparse.urlparse(url)
  238.         embed_url = compat_urlparse.urlunparse(
  239.             parsed_url._replace(path=parsed_url.path.rsplit('/', 1)[0] + '/embed'))
  241.         webpage, urlh = self._download_webpage_handle(embed_url, video_id)
  242.         new_url = urlh.geturl()
  243.         ign_url = compat_parse_qs(
  244.             compat_urlparse.urlparse(new_url).query).get('url', [None])[-1]
  245.         if ign_url:
  246.             return self.url_result(ign_url, IGNIE.ie_key())
  247.         video = self._search_regex(r'(<div\b[^>]+\bdata-video-id\s*=\s*[^>]+>)', webpage, 'video element', fatal=False)
  248.         if not video:
  249.             if new_url == url:
  250.                 raise ExtractorError('Redirect loop: ' + url)
  251.             return self.url_result(new_url)
  252.         video = extract_attributes(video)
  253.         video_data = video.get('data-settings') or '{}'
  254.         video_data = self._parse_json(video_data, video_id)['video']
  255.         info = self._extract_video_info(video_data)
  257.         return merge_dicts({
  258.             'display_id': video_id,
  259.         }, info)
  262. class IGNArticleIE(IGNBaseIE):
  263.     _VALID_URL = r'https?://.+?\.ign\.com/(?:articles(?:/\d{4}/\d{2}/\d{2})?|(?:[a-z]{2}/)?(?:[\w-]+/)*?feature/\d+)/(?P<id>[^/?&#]+)'
  264.     _PAGE_TYPE = 'article'
  265.     _TESTS = [{
  266.         'url': 'http://me.ign.com/en/feature/15775/100-little-things-in-gta-5-that-will-blow-your-mind',
  267.         'info_dict': {
  268.             'id': '72113',
  269.             'title': '100 Little Things in GTA 5 That Will Blow Your Mind',
  270.         },
  271.         'playlist': [
  272.             {
  273.                 'info_dict': {
  274.                     'id': '5ebbd138523268b93c9141af17bec937',
  275.                     'ext': 'mp4',
  276.                     'title': 'Grand Theft Auto V Video Review',
  277.                     'description': 'Rockstar drops the mic on this generation of games. Watch our review of the masterly Grand Theft Auto V.',
  278.                     'timestamp': 1379339880,
  279.                     'upload_date': '20130916',
  280.                 },
  281.             },
  282.             {
  283.                 'info_dict': {
  284.                     'id': '638672ee848ae4ff108df2a296418ee2',
  285.                     'ext': 'mp4',
  286.                     'title': 'GTA 5 In Slow Motion',
  287.                     'description': 'The twisted beauty of GTA 5 in stunning slow motion.',
  288.                     'timestamp': 1386878820,
  289.                     'upload_date': '20131212',
  290.                 },
  291.             },
  292.         ],
  293.         'params': {
  294.             'skip_download': True,
  295.         },
  296.         'expected_warnings': ['Backend fetch failed'],
  297.     }, {
  298.         'url': 'http://www.ign.com/articles/2014/08/15/rewind-theater-wild-trailer-gamescom-2014?watch',
  299.         'info_dict': {
  300.             'id': '53ee806780a81ec46e0790f8',
  301.             'title': 'Rewind Theater - Wild Trailer Gamescom 2014',
  302.         },
  303.         'playlist_count': 1,
  304.         'expected_warnings': ['Backend fetch failed'],
  305.     }, {
  306.         # videoId pattern
  307.         'url': 'http://www.ign.com/articles/2017/06/08/new-ducktales-short-donalds-birthday-doesnt-go-as-planned',
  308.         'only_matching': True,
  309.     }, {
  310.         # Youtube embed
  311.         'url': 'https://www.ign.com/articles/2021-mvp-named-in-puppy-bowl-xvii',
  312.         'only_matching': True,
  313.     }, {
  314.         # IMDB embed
  315.         'url': 'https://www.ign.com/articles/2014/08/07/sons-of-anarchy-final-season-trailer',
  316.         'only_matching': True,
  317.     }, {
  318.         # Facebook embed
  319.         'url': 'https://www.ign.com/articles/2017/09/20/marvels-the-punisher-watch-the-new-trailer-for-the-netflix-series',
  320.         'only_matching': True,
  321.     }, {
  322.         # Brightcove embed
  323.         'url': 'https://www.ign.com/articles/2016/01/16/supergirl-goes-flying-with-martian-manhunter-in-new-clip',
  324.         'only_matching': True,
  325.     }]
  327.     def _checked_call_api(self, slug):
  328.         try:
  329.             return self._call_api(slug)
  330.         except ExtractorError as e:
  331.             if isinstance(e.cause, compat_HTTPError):
  332.                 e.cause.args = e.cause.args or [
  333.                     e.cause.geturl(), e.cause.getcode(), e.cause.reason]
  334.                 if e.cause.code == 404:
  335.                     raise ExtractorError(
  336.                         'Content not found: expired?', cause=e.cause,
  337.                         expected=True)
  338.                 elif e.cause.code == 503:
  339.                     self.report_warning(error_to_compat_str(e.cause))
  340.                     return
  341.             raise
  343.     def _search_nextjs_data(self, webpage, video_id, **kw):
  344.         return self._parse_json(
  345.             self._search_regex(
  346.                 r'(?s)<script[^>]+id=[\'"]__NEXT_DATA__[\'"][^>]*>([^<]+)</script>',
  347.                 webpage, 'next.js data', **kw),
  348.             video_id, **kw)
  350.     def _real_extract(self, url):
  351.         display_id = self._match_id(url)
  352.         article = self._checked_call_api(display_id)
  354.         if article:
  355.             # obsolete ?
  356.             def entries():
  357.                 media_url = traverse_obj(
  358.                     article, ('mediaRelations', 0, 'media', 'metadata', 'url'),
  359.                     expected_type=url_or_none)
  360.                 if media_url:
  361.                     yield self.url_result(media_url, IGNIE.ie_key())
  362.                 for content in (article.get('content') or []):
  363.                     for video_url in re.findall(r'(?:\[(?:ignvideo\s+url|youtube\s+clip_id)|<iframe[^>]+src)="([^"]+)"', content):
  364.                         if url_or_none(video_url):
  365.                             yield self.url_result(video_url)
  367.             return self.playlist_result(
  368.                 entries(), article.get('articleId'),
  369.                 traverse_obj(
  370.                     article, ('metadata', 'headline'),
  371.                     expected_type=lambda x: x.strip() or None))
  373.         webpage = self._download_webpage(url, display_id)
  375.         playlist_id = self._html_search_meta('dable:item_id', webpage, default=None)
  376.         if playlist_id:
  378.             def entries():
  379.                 for m in re.finditer(
  380.                         r'''(?s)<object\b[^>]+\bclass\s*=\s*("|')ign-videoplayer\1[^>]*>(?P<params>.+?)</object''',
  381.                         webpage):
  382.                     flashvars = self._search_regex(
  383.                         r'''(<param\b[^>]+\bname\s*=\s*("|')flashvars\2[^>]*>)''',
  384.                         m.group('params'), 'flashvars', default='')
  385.                     flashvars = compat_parse_qs(extract_attributes(flashvars).get('value') or '')
  386.                     v_url = url_or_none((flashvars.get('url') or [None])[-1])
  387.                     if v_url:
  388.                         yield self.url_result(v_url)
  389.         else:
  390.             playlist_id = self._search_regex(
  391.                 r'''\bdata-post-id\s*=\s*("|')(?P<id>[\da-f]+)\1''',
  392.                 webpage, 'id', group='id', default=None)
  394.             nextjs_data = self._search_nextjs_data(webpage, display_id)
  396.             def entries():
  397.                 for player in traverse_obj(
  398.                         nextjs_data,
  399.                         ('props', 'apolloState', 'ROOT_QUERY', lambda k, _: k.startswith('videoPlayerProps('), '__ref')):
  400.                     # skip promo links (which may not always be served, eg GH CI servers)
  401.                     if traverse_obj(nextjs_data,
  402.                                     ('props', 'apolloState', player.replace('PlayerProps', 'ModernContent')),
  403.                                     expected_type=dict):
  404.                         continue
  405.                     video = traverse_obj(nextjs_data, ('props', 'apolloState', player), expected_type=dict) or {}
  406.                     info = self._extract_video_info(video, fatal=False)
  407.                     if info:
  408.                         yield merge_dicts({
  409.                             'display_id': display_id,
  410.                         }, info)
  412.         return self.playlist_result(
  413.             entries(), playlist_id or display_id,
  414.             re.sub(r'\s+-\s+IGN\s*$', '', self._og_search_title(webpage, default='')) or None)