[tf1] improve extraction(closes #27980)(closes #28040)

youtube_dl/extractor/tf1.py

@@ -1,92 +1,87 @@
 # coding: utf-8
 from __future__ import unicode_literals
 
+import json
+import re
+
 from .common import InfoExtractor
-from ..compat import compat_str
+from ..utils import (
+    int_or_none,
+    parse_iso8601,
+    try_get,
+)
 
 
 class TF1IE(InfoExtractor):
-    """TF1 uses the wat.tv player."""
-    _VALID_URL = r'https?://(?:(?:videos|www|lci)\.tf1|(?:www\.)?(?:tfou|ushuaiatv|histoire|tvbreizh))\.fr/(?:[^/]+/)*(?P<id>[^/?#.]+)'
+    _VALID_URL = r'https?://(?:www\.)?tf1\.fr/[^/]+/(?P<program_slug>[^/]+)/videos/(?P<id>[^/?&#]+)\.html'
     _TESTS = [{
-        'url': 'http://videos.tf1.fr/auto-moto/citroen-grand-c4-picasso-2013-presentation-officielle-8062060.html',
-        'info_dict': {
-            'id': '10635995',
-            'ext': 'mp4',
-            'title': 'Citroën Grand C4 Picasso 2013 : présentation officielle',
-            'description': 'Vidéo officielle du nouveau Citroën Grand C4 Picasso, lancé à l\'automne 2013.',
-        },
-        'params': {
-            # Sometimes wat serves the whole file with the --test option
-            'skip_download': True,
-        },
-        'expected_warnings': ['HTTP Error 404'],
-    }, {
-        'url': 'http://www.tfou.fr/chuggington/videos/le-grand-mysterioso-chuggington-7085291-739.html',
+        'url': 'https://www.tf1.fr/tmc/quotidien-avec-yann-barthes/videos/quotidien-premiere-partie-11-juin-2019.html',
         'info_dict': {
-            'id': 'le-grand-mysterioso-chuggington-7085291-739',
+            'id': '13641379',
             'ext': 'mp4',
-            'title': 'Le grand Mystérioso - Chuggington',
-            'description': 'Le grand Mystérioso - Emery rêve qu\'un article lui soit consacré dans le journal.',
-            'upload_date': '20150103',
+            'title': 'md5:f392bc52245dc5ad43771650c96fb620',
+            'description': 'md5:a02cdb217141fb2d469d6216339b052f',
+            'upload_date': '20190611',
+            'timestamp': 1560273989,
+            'duration': 1738,
+            'series': 'Quotidien avec Yann Barthès',
+            'tags': ['intégrale', 'quotidien', 'Replay'],
         },
         'params': {
             # Sometimes wat serves the whole file with the --test option
             'skip_download': True,
+            'format': 'bestvideo',
         },
-        'skip': 'HTTP Error 410: Gone',
     }, {
         'url': 'http://www.tf1.fr/tf1/koh-lanta/videos/replay-koh-lanta-22-mai-2015.html',
         'only_matching': True,
-    }, {
-        'url': 'http://lci.tf1.fr/sept-a-huit/videos/sept-a-huit-du-24-mai-2015-8611550.html',
-        'only_matching': True,
     }, {
         'url': 'http://www.tf1.fr/hd1/documentaire/videos/mylene-farmer-d-une-icone.html',
         'only_matching': True,
-    }, {
-        'url': 'https://www.tf1.fr/tmc/quotidien-avec-yann-barthes/videos/quotidien-premiere-partie-11-juin-2019.html',
-        'info_dict': {
-            'id': '13641379',
-            'ext': 'mp4',
-            'title': 'md5:f392bc52245dc5ad43771650c96fb620',
-            'description': 'md5:44bc54f0a21322f5b91d68e76a544eae',
-            'upload_date': '20190611',
-        },
-        'params': {
-            # Sometimes wat serves the whole file with the --test option
-            'skip_download': True,
-        },
     }]
 
     def _real_extract(self, url):
-        video_id = self._match_id(url)
-
-        webpage = self._download_webpage(url, video_id)
-
-        wat_id = None
+        program_slug, slug = re.match(self._VALID_URL, url).groups()
+        video = self._download_json(
+            'https://www.tf1.fr/graphql/web', slug, query={
+                'id': '9b80783950b85247541dd1d851f9cc7fa36574af015621f853ab111a679ce26f',
+                'variables': json.dumps({
+                    'programSlug': program_slug,
+                    'slug': slug,
+                })
+            })['data']['videoBySlug']
+        wat_id = video['streamId']
 
-        data = self._parse_json(
-            self._search_regex(
-                r'__APOLLO_STATE__\s*=\s*({.+?})\s*(?:;|</script>)', webpage,
-                'data', default='{}'), video_id, fatal=False)
+        tags = []
+        for tag in (video.get('tags') or []):
+            label = tag.get('label')
+            if not label:
+                continue
+            tags.append(label)
 
-        if data:
-            try:
-                wat_id = next(
-                    video.get('streamId')
-                    for key, video in data.items()
-                    if isinstance(video, dict)
-                    and video.get('slug') == video_id)
-                if not isinstance(wat_id, compat_str) or not wat_id.isdigit():
-                    wat_id = None
-            except StopIteration:
-                pass
+        decoration = video.get('decoration') or {}
 
-        if not wat_id:
-            wat_id = self._html_search_regex(
-                (r'(["\'])(?:https?:)?//www\.wat\.tv/embedframe/.*?(?P<id>\d{8})\1',
-                 r'(["\']?)streamId\1\s*:\s*(["\']?)(?P<id>\d+)\2'),
-                webpage, 'wat id', group='id')
+        thumbnails = []
+        for source in (try_get(decoration, lambda x: x['image']['sources'], list) or []):
+            source_url = source.get('url')
+            if not source_url:
+                continue
+            thumbnails.append({
+                'url': source_url,
+                'width': int_or_none(source.get('width')),
+            })
 
-        return self.url_result('wat:%s' % wat_id, 'Wat')
+        return {
+            '_type': 'url_transparent',
+            'id': wat_id,
+            'url': 'wat:' + wat_id,
+            'title': video.get('title'),
+            'thumbnails': thumbnails,
+            'description': decoration.get('description'),
+            'timestamp': parse_iso8601(video.get('date')),
+            'duration': int_or_none(try_get(video, lambda x: x['publicPlayingInfos']['duration'])),
+            'tags': tags,
+            'series': decoration.get('programLabel'),
+            'season_number': int_or_none(video.get('season')),
+            'episode_number': int_or_none(video.get('episode')),
+        }

youtube_dl/extractor/wat.py

@@ -4,9 +4,10 @@ from __future__ import unicode_literals
 from .common import InfoExtractor
 from ..compat import compat_str
 from ..utils import (
-    unified_strdate,
-    HEADRequest,
+    ExtractorError,
     int_or_none,
+    try_get,
+    unified_strdate,
 )
 
 
@@ -29,6 +30,7 @@ class WatIE(InfoExtractor):
                 'skip_download': True,
             },
             'expected_warnings': ['HTTP Error 404'],
+            'skip': 'This content is no longer available',
         },
         {
             'url': 'http://www.wat.tv/video/gregory-lemarchal-voix-ange-6z1v7_6ygkj_.html',
@@ -40,8 +42,10 @@ class WatIE(InfoExtractor):
                 'upload_date': '20140816',
             },
             'expected_warnings': ["Ce contenu n'est pas disponible pour l'instant."],
+            'skip': 'This content is no longer available',
         },
     ]
+    _GEO_BYPASS = False
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
@@ -49,71 +53,52 @@ class WatIE(InfoExtractor):
 
         # 'contentv4' is used in the website, but it also returns the related
         # videos, we don't need them
+        # video_data = self._download_json(
+        #     'http://www.wat.tv/interface/contentv4s/' + video_id, video_id)
         video_data = self._download_json(
-            'http://www.wat.tv/interface/contentv4s/' + video_id, video_id)
+            'https://mediainfo.tf1.fr/mediainfocombo/' + video_id,
+            video_id, query={'context': 'MYTF1'})
         video_info = video_data['media']
 
         error_desc = video_info.get('error_desc')
         if error_desc:
-            self.report_warning(
-                '%s returned error: %s' % (self.IE_NAME, error_desc))
-
-        chapters = video_info['chapters']
-        if chapters:
-            first_chapter = chapters[0]
+            if video_info.get('error_code') == 'GEOBLOCKED':
+                self.raise_geo_restricted(error_desc, video_info.get('geoList'))
+            raise ExtractorError(error_desc)
 
-            def video_id_for_chapter(chapter):
-                return chapter['tc_start'].split('-')[0]
+        title = video_info['title']
 
-            if video_id_for_chapter(first_chapter) != video_id:
-                self.to_screen('Multipart video detected')
-                entries = [self.url_result('wat:%s' % video_id_for_chapter(chapter)) for chapter in chapters]
-                return self.playlist_result(entries, video_id, video_info['title'])
-            # Otherwise we can continue and extract just one part, we have to use
-            # the video id for getting the video url
-        else:
-            first_chapter = video_info
+        formats = []
 
-        title = first_chapter['title']
+        def extract_formats(manifest_urls):
+            for f, f_url in manifest_urls.items():
+                if not f_url:
+                    continue
+                if f in ('dash', 'mpd'):
+                    formats.extend(self._extract_mpd_formats(
+                        f_url.replace('://das-q1.tf1.fr/', '://das-q1-ssl.tf1.fr/'),
+                        video_id, mpd_id='dash', fatal=False))
+                elif f == 'hls':
+                    formats.extend(self._extract_m3u8_formats(
+                        f_url, video_id, 'mp4',
+                        'm3u8_native', m3u8_id='hls', fatal=False))
 
-        def extract_url(path_template, url_type):
-            req_url = 'http://www.wat.tv/get/%s' % (path_template % video_id)
-            head = self._request_webpage(HEADRequest(req_url), video_id, 'Extracting %s url' % url_type, fatal=False)
-            if head:
-                red_url = head.geturl()
-                if req_url != red_url:
-                    return red_url
-            return None
+        delivery = video_data.get('delivery') or {}
+        extract_formats({delivery.get('format'): delivery.get('url')})
+        if not formats:
+            manifest_urls = self._download_json(
+                'http://www.wat.tv/get/webhtml/' + video_id, video_id, fatal=False)
+            if manifest_urls:
+                extract_formats(manifest_urls)
 
-        formats = []
-        manifest_urls = self._download_json(
-            'http://www.wat.tv/get/webhtml/' + video_id, video_id)
-        m3u8_url = manifest_urls.get('hls')
-        if m3u8_url:
-            formats.extend(self._extract_m3u8_formats(
-                m3u8_url, video_id, 'mp4',
-                'm3u8_native', m3u8_id='hls', fatal=False))
-        mpd_url = manifest_urls.get('mpd')
-        if mpd_url:
-            formats.extend(self._extract_mpd_formats(
-                mpd_url.replace('://das-q1.tf1.fr/', '://das-q1-ssl.tf1.fr/'),
-                video_id, mpd_id='dash', fatal=False))
         self._sort_formats(formats)
 
-        date_diffusion = first_chapter.get('date_diffusion') or video_data.get('configv4', {}).get('estatS4')
-        upload_date = unified_strdate(date_diffusion) if date_diffusion else None
-        duration = None
-        files = video_info['files']
-        if files:
-            duration = int_or_none(files[0].get('duration'))
-
         return {
             'id': video_id,
             'title': title,
-            'thumbnail': first_chapter.get('preview'),
-            'description': first_chapter.get('description'),
-            'view_count': int_or_none(video_info.get('views')),
-            'upload_date': upload_date,
-            'duration': duration,
+            'thumbnail': video_info.get('preview'),
+            'upload_date': unified_strdate(try_get(
+                video_data, lambda x: x['mediametrie']['chapters'][0]['estatS4'])),
+            'duration': int_or_none(video_info.get('duration')),
             'formats': formats,
         }