[ign] fix extraction(closes #24771)

youtube_dl/extractor/extractors.py

@@ -470,8 +470,8 @@ from .hungama import (
 from .hypem import HypemIE
 from .ign import (
     IGNIE,
-    OneUPIE,
+    IGNVideoIE,
-    PCMagIE,
+    IGNArticleIE,
 )
 from .iheart import (
     IHeartRadioIE,

youtube_dl/extractor/ign.py

@@ -3,28 +3,39 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
+from ..compat import (
+    compat_parse_qs,
+    compat_urllib_parse_urlparse,
+)
 from ..utils import (
+    HEADRequest,
+    determine_ext,
     int_or_none,
     parse_iso8601,
+    strip_or_none,
+    try_get,
 )
 
 
-class IGNIE(InfoExtractor):
+class IGNBaseIE(InfoExtractor):
+    def _call_api(self, slug):
+        return self._download_json(
+            'http://apis.ign.com/{0}/v3/{0}s/slug/{1}'.format(self._PAGE_TYPE, slug), slug)
+
+
+class IGNIE(IGNBaseIE):
     """
     Extractor for some of the IGN sites, like www.ign.com, es.ign.com de.ign.com.
     Some videos of it.ign.com are also supported
     """
 
-    _VALID_URL = r'https?://.+?\.ign\.com/(?:[^/]+/)?(?P<type>videos|show_videos|articles|feature|(?:[^/]+/\d+/video))(/.+)?/(?P<name_or_id>.+)'
+    _VALID_URL = r'https?://(?:.+?\.ign|www\.pcmag)\.com/videos/(?:\d{4}/\d{2}/\d{2}/)?(?P<id>[^/?&#]+)'
     IE_NAME = 'ign.com'
+    _PAGE_TYPE = 'video'
 
-    _API_URL_TEMPLATE = 'http://apis.ign.com/video/v3/videos/%s'
+    _TESTS = [{
-    _EMBED_RE = r'<iframe[^>]+?["\']((?:https?:)?//.+?\.ign\.com.+?/embed.+?)["\']'
-
-    _TESTS = [
-        {
         'url': 'http://www.ign.com/videos/2013/06/05/the-last-of-us-review',
-            'md5': 'febda82c4bafecd2d44b6e1a18a595f8',
+        'md5': 'd2e1586d9987d40fad7867bf96a018ea',
         'info_dict': {
             'id': '8f862beef863986b2785559b9e1aa599',
             'ext': 'mp4',
@@ -32,13 +43,147 @@ class IGNIE(InfoExtractor):
             'description': 'md5:c8946d4260a4d43a00d5ae8ed998870c',
             'timestamp': 1370440800,
             'upload_date': '20130605',
-                'uploader_id': 'cberidon@ign.com',
+            'tags': 'count:9',
         }
-        },
+    }, {
-        {
+        'url': 'http://www.pcmag.com/videos/2015/01/06/010615-whats-new-now-is-gogo-snooping-on-your-data',
+        'md5': 'f1581a6fe8c5121be5b807684aeac3f6',
+        'info_dict': {
+            'id': 'ee10d774b508c9b8ec07e763b9125b91',
+            'ext': 'mp4',
+            'title': 'What\'s New Now: Is GoGo Snooping on Your Data?',
+            'description': 'md5:817a20299de610bd56f13175386da6fa',
+            'timestamp': 1420571160,
+            'upload_date': '20150106',
+            'tags': 'count:4',
+        }
+    }, {
+        'url': 'https://www.ign.com/videos/is-a-resident-evil-4-remake-on-the-way-ign-daily-fix',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        display_id = self._match_id(url)
+        video = self._call_api(display_id)
+        video_id = video['videoId']
+        metadata = video['metadata']
+        title = metadata.get('longTitle') or metadata.get('title') or metadata['name']
+
+        formats = []
+        refs = video.get('refs') or {}
+
+        m3u8_url = refs.get('m3uUrl')
+        if m3u8_url:
+            formats.extend(self._extract_m3u8_formats(
+                m3u8_url, video_id, 'mp4', 'm3u8_native',
+                m3u8_id='hls', fatal=False))
+
+        f4m_url = refs.get('f4mUrl')
+        if f4m_url:
+            formats.extend(self._extract_f4m_formats(
+                f4m_url, video_id, f4m_id='hds', fatal=False))
+
+        for asset in (video.get('assets') or []):
+            asset_url = asset.get('url')
+            if not asset_url:
+                continue
+            formats.append({
+                'url': asset_url,
+                'tbr': int_or_none(asset.get('bitrate'), 1000),
+                'fps': int_or_none(asset.get('frame_rate')),
+                'height': int_or_none(asset.get('height')),
+                'width': int_or_none(asset.get('width')),
+            })
+
+        mezzanine_url = try_get(video, lambda x: x['system']['mezzanineUrl'])
+        if mezzanine_url:
+            formats.append({
+                'ext': determine_ext(mezzanine_url, 'mp4'),
+                'format_id': 'mezzanine',
+                'preference': 1,
+                'url': mezzanine_url,
+            })
+
+        self._sort_formats(formats)
+
+        thumbnails = []
+        for thumbnail in (video.get('thumbnails') or []):
+            thumbnail_url = thumbnail.get('url')
+            if not thumbnail_url:
+                continue
+            thumbnails.append({
+                'url': thumbnail_url,
+            })
+
+        tags = []
+        for tag in (video.get('tags') or []):
+            display_name = tag.get('displayName')
+            if not display_name:
+                continue
+            tags.append(display_name)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': strip_or_none(metadata.get('description')),
+            'timestamp': parse_iso8601(metadata.get('publishDate')),
+            'duration': int_or_none(metadata.get('duration')),
+            'display_id': display_id,
+            'thumbnails': thumbnails,
+            'formats': formats,
+            'tags': tags,
+        }
+
+
+class IGNVideoIE(InfoExtractor):
+    _VALID_URL = r'https?://.+?\.ign\.com/(?:[a-z]{2}/)?[^/]+/(?P<id>\d+)/(?:video|trailer)/'
+    _TESTS = [{
+        'url': 'http://me.ign.com/en/videos/112203/video/how-hitman-aims-to-be-different-than-every-other-s',
+        'md5': 'dd9aca7ed2657c4e118d8b261e5e9de1',
+        'info_dict': {
+            'id': 'e9be7ea899a9bbfc0674accc22a36cc8',
+            'ext': 'mp4',
+            'title': 'How Hitman Aims to Be Different Than Every Other Stealth Game - NYCC 2015',
+            'description': 'Taking out assassination targets in Hitman has never been more stylish.',
+            'timestamp': 1444665600,
+            'upload_date': '20151012',
+        }
+    }, {
+        'url': 'http://me.ign.com/ar/angry-birds-2/106533/video/lrd-ldyy-lwl-lfylm-angry-birds',
+        'only_matching': True,
+    }, {
+        # Youtube embed
+        'url': 'https://me.ign.com/ar/ratchet-clank-rift-apart/144327/trailer/embed',
+        'only_matching': True,
+    }, {
+        # Twitter embed
+        'url': 'http://adria.ign.com/sherlock-season-4/9687/trailer/embed',
+        'only_matching': True,
+    }, {
+        # Vimeo embed
+        'url': 'https://kr.ign.com/bic-2018/3307/trailer/embed',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        req = HEADRequest(url.rsplit('/', 1)[0] + '/embed')
+        url = self._request_webpage(req, video_id).geturl()
+        ign_url = compat_parse_qs(
+            compat_urllib_parse_urlparse(url).query).get('url', [None])[0]
+        if ign_url:
+            return self.url_result(ign_url, IGNIE.ie_key())
+        return self.url_result(url)
+
+
+class IGNArticleIE(IGNBaseIE):
+    _VALID_URL = r'https?://.+?\.ign\.com/(?:articles(?:/\d{4}/\d{2}/\d{2})?|(?:[a-z]{2}/)?feature/\d+)/(?P<id>[^/?&#]+)'
+    _PAGE_TYPE = 'article'
+    _TESTS = [{
         'url': 'http://me.ign.com/en/feature/15775/100-little-things-in-gta-5-that-will-blow-your-mind',
         'info_dict': {
-                'id': '100-little-things-in-gta-5-that-will-blow-your-mind',
+            'id': '524497489e4e8ff5848ece34',
+            'title': '100 Little Things in GTA 5 That Will Blow Your Mind',
         },
         'playlist': [
             {
@@ -49,7 +194,6 @@ class IGNIE(InfoExtractor):
                     'description': 'Rockstar drops the mic on this generation of games. Watch our review of the masterly Grand Theft Auto V.',
                     'timestamp': 1379339880,
                     'upload_date': '20130916',
-                        'uploader_id': 'danieljkrupa@gmail.com',
                 },
             },
             {
@@ -60,173 +204,54 @@ class IGNIE(InfoExtractor):
                     'description': 'The twisted beauty of GTA 5 in stunning slow motion.',
                     'timestamp': 1386878820,
                     'upload_date': '20131212',
-                        'uploader_id': 'togilvie@ign.com',
                 },
             },
         ],
         'params': {
+            'playlist_items': '2-3',
             'skip_download': True,
         },
-        },
+    }, {
-        {
         'url': 'http://www.ign.com/articles/2014/08/15/rewind-theater-wild-trailer-gamescom-2014?watch',
-            'md5': '618fedb9c901fd086f6f093564ef8558',
         'info_dict': {
-                'id': '078fdd005f6d3c02f63d795faa1b984f',
+            'id': '53ee806780a81ec46e0790f8',
-                'ext': 'mp4',
             'title': 'Rewind Theater - Wild Trailer Gamescom 2014',
-                'description': 'Brian and Jared explore Michel Ancel\'s captivating new preview.',
-                'timestamp': 1408047180,
-                'upload_date': '20140814',
-                'uploader_id': 'jamesduggan1990@gmail.com',
         },
-        },
+        'playlist_count': 2,
-        {
+    }, {
-            'url': 'http://me.ign.com/en/videos/112203/video/how-hitman-aims-to-be-different-than-every-other-s',
-            'only_matching': True,
-        },
-        {
-            'url': 'http://me.ign.com/ar/angry-birds-2/106533/video/lrd-ldyy-lwl-lfylm-angry-birds',
-            'only_matching': True,
-        },
-        {
         # videoId pattern
         'url': 'http://www.ign.com/articles/2017/06/08/new-ducktales-short-donalds-birthday-doesnt-go-as-planned',
         'only_matching': True,
-        },
-    ]
-
-    def _find_video_id(self, webpage):
-        res_id = [
-            r'"video_id"\s*:\s*"(.*?)"',
-            r'class="hero-poster[^"]*?"[^>]*id="(.+?)"',
-            r'data-video-id="(.+?)"',
-            r'<object id="vid_(.+?)"',
-            r'<meta name="og:image" content=".*/(.+?)-(.+?)/.+.jpg"',
-            r'videoId&quot;\s*:\s*&quot;(.+?)&quot;',
-            r'videoId["\']\s*:\s*["\']([^"\']+?)["\']',
-        ]
-        return self._search_regex(res_id, webpage, 'video id', default=None)
-
-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        name_or_id = mobj.group('name_or_id')
-        page_type = mobj.group('type')
-        webpage = self._download_webpage(url, name_or_id)
-        if page_type != 'video':
-            multiple_urls = re.findall(
-                r'<param name="flashvars"[^>]*value="[^"]*?url=(https?://www\.ign\.com/videos/.*?)["&]',
-                webpage)
-            if multiple_urls:
-                entries = [self.url_result(u, ie='IGN') for u in multiple_urls]
-                return {
-                    '_type': 'playlist',
-                    'id': name_or_id,
-                    'entries': entries,
-                }
-
-        video_id = self._find_video_id(webpage)
-        if not video_id:
-            return self.url_result(self._search_regex(
-                self._EMBED_RE, webpage, 'embed url'))
-        return self._get_video_info(video_id)
-
-    def _get_video_info(self, video_id):
-        api_data = self._download_json(
-            self._API_URL_TEMPLATE % video_id, video_id)
-
-        formats = []
-        m3u8_url = api_data['refs'].get('m3uUrl')
-        if m3u8_url:
-            formats.extend(self._extract_m3u8_formats(
-                m3u8_url, video_id, 'mp4', 'm3u8_native',
-                m3u8_id='hls', fatal=False))
-        f4m_url = api_data['refs'].get('f4mUrl')
-        if f4m_url:
-            formats.extend(self._extract_f4m_formats(
-                f4m_url, video_id, f4m_id='hds', fatal=False))
-        for asset in api_data['assets']:
-            formats.append({
-                'url': asset['url'],
-                'tbr': asset.get('actual_bitrate_kbps'),
-                'fps': asset.get('frame_rate'),
-                'height': int_or_none(asset.get('height')),
-                'width': int_or_none(asset.get('width')),
-            })
-        self._sort_formats(formats)
-
-        thumbnails = [{
-            'url': thumbnail['url']
-        } for thumbnail in api_data.get('thumbnails', [])]
-
-        metadata = api_data['metadata']
-
-        return {
-            'id': api_data.get('videoId') or video_id,
-            'title': metadata.get('longTitle') or metadata.get('name') or metadata.get['title'],
-            'description': metadata.get('description'),
-            'timestamp': parse_iso8601(metadata.get('publishDate')),
-            'duration': int_or_none(metadata.get('duration')),
-            'display_id': metadata.get('slug') or video_id,
-            'uploader_id': metadata.get('creator'),
-            'thumbnails': thumbnails,
-            'formats': formats,
-        }
-
-
-class OneUPIE(IGNIE):
-    _VALID_URL = r'https?://gamevideos\.1up\.com/(?P<type>video)/id/(?P<name_or_id>.+)\.html'
-    IE_NAME = '1up.com'
-
-    _TESTS = [{
-        'url': 'http://gamevideos.1up.com/video/id/34976.html',
-        'md5': 'c9cc69e07acb675c31a16719f909e347',
-        'info_dict': {
-            'id': '34976',
-            'ext': 'mp4',
-            'title': 'Sniper Elite V2 - Trailer',
-            'description': 'md5:bf0516c5ee32a3217aa703e9b1bc7826',
-            'timestamp': 1313099220,
-            'upload_date': '20110811',
-            'uploader_id': 'IGN',
-        }
-    }]
-
-    def _real_extract(self, url):
-        mobj = re.match(self._VALID_URL, url)
-        result = super(OneUPIE, self)._real_extract(url)
-        result['id'] = mobj.group('name_or_id')
-        return result
-
-
-class PCMagIE(IGNIE):
-    _VALID_URL = r'https?://(?:www\.)?pcmag\.com/(?P<type>videos|article2)(/.+)?/(?P<name_or_id>.+)'
-    IE_NAME = 'pcmag'
-
-    _EMBED_RE = r'iframe\.setAttribute\("src",\s*__util.objToUrlString\("http://widgets\.ign\.com/video/embed/content\.html?[^"]*url=([^"]+)["&]'
-
-    _TESTS = [{
-        'url': 'http://www.pcmag.com/videos/2015/01/06/010615-whats-new-now-is-gogo-snooping-on-your-data',
-        'md5': '212d6154fd0361a2781075f1febbe9ad',
-        'info_dict': {
-            'id': 'ee10d774b508c9b8ec07e763b9125b91',
-            'ext': 'mp4',
-            'title': '010615_What\'s New Now: Is GoGo Snooping on Your Data?',
-            'description': 'md5:a7071ae64d2f68cc821c729d4ded6bb3',
-            'timestamp': 1420571160,
-            'upload_date': '20150106',
-            'uploader_id': 'cozzipix@gmail.com',
-        }
     }, {
-        'url': 'http://www.pcmag.com/article2/0,2817,2470156,00.asp',
+        # Youtube embed
-        'md5': '94130c1ca07ba0adb6088350681f16c1',
+        'url': 'https://www.ign.com/articles/2021-mvp-named-in-puppy-bowl-xvii',
-        'info_dict': {
+        'only_matching': True,
-            'id': '042e560ba94823d43afcb12ddf7142ca',
+    }, {
-            'ext': 'mp4',
+        # IMDB embed
-            'title': 'HTC\'s Weird New Re Camera - What\'s New Now',
+        'url': 'https://www.ign.com/articles/2014/08/07/sons-of-anarchy-final-season-trailer',
-            'description': 'md5:53433c45df96d2ea5d0fda18be2ca908',
+        'only_matching': True,
-            'timestamp': 1412953920,
+    }, {
-            'upload_date': '20141010',
+        # Facebook embed
-            'uploader_id': 'chris_snyder@pcmag.com',
+        'url': 'https://www.ign.com/articles/2017/09/20/marvels-the-punisher-watch-the-new-trailer-for-the-netflix-series',
-        }
+        'only_matching': True,
+    }, {
+        # Brightcove embed
+        'url': 'https://www.ign.com/articles/2016/01/16/supergirl-goes-flying-with-martian-manhunter-in-new-clip',
+        'only_matching': True,
     }]
+
+    def _real_extract(self, url):
+        display_id = self._match_id(url)
+        article = self._call_api(display_id)
+
+        def entries():
+            media_url = try_get(article, lambda x: x['mediaRelations'][0]['media']['metadata']['url'])
+            if media_url:
+                yield self.url_result(media_url, IGNIE.ie_key())
+            for content in (article.get('content') or []):
+                for video_url in re.findall(r'(?:\[(?:ignvideo\s+url|youtube\s+clip_id)|<iframe[^>]+src)="([^"]+)"', content):
+                    yield self.url_result(video_url)
+
+        return self.playlist_result(
+            entries(), article.get('articleId'),
+            strip_or_none(try_get(article, lambda x: x['metadata']['headline'])))