[arte] Rework extractors
authorSergey M․ <dstftw@gmail.com>
Wed, 18 Nov 2020 22:02:04 +0000 (05:02 +0700)
committerSergey M․ <dstftw@gmail.com>
Wed, 18 Nov 2020 22:02:04 +0000 (05:02 +0700)
* Reimplement embed and playlist extractors to delegate to the single entrypoint artetv extractor
  Beware reluctant download archive extractor keys breakage.
* Improve embeds detection (closes #27057)
- Remove obsolete code

youtube_dl/extractor/arte.py
youtube_dl/extractor/extractors.py
youtube_dl/extractor/generic.py

index b80467548099a29e51b9c00f614e4c399c73c7ef..03abdbfafa7cfb5c5aedf289cbe7ce0bb6d7ab35 100644 (file)
@@ -4,7 +4,10 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..compat import compat_str
+from ..compat import (
+    compat_str,
+    compat_urlparse,
+)
 from ..utils import (
     ExtractorError,
     int_or_none,
@@ -14,14 +17,44 @@ from ..utils import (
     url_or_none,
 )
 
-# There are different sources of video in arte.tv, the extraction process
-# is different for each one. The videos usually expire in 7 days, so we can't
-# add tests.
-
 
 class ArteTVBaseIE(InfoExtractor):
-    def _extract_from_json_url(self, json_url, video_id, lang, title=None):
-        info = self._download_json(json_url, video_id)
+    _ARTE_LANGUAGES = 'fr|de|en|es|it|pl'
+    _API_BASE = 'https://api.arte.tv/api/player/v1'
+
+
+class ArteTVIE(ArteTVBaseIE):
+    _VALID_URL = r'''(?x)
+                    https?://
+                        (?:
+                            (?:www\.)?arte\.tv/(?P<lang>%(langs)s)/videos|
+                            api\.arte\.tv/api/player/v\d+/config/(?P<lang_2>%(langs)s)
+                        )
+                        /(?P<id>\d{6}-\d{3}-[AF])
+                    ''' % {'langs': ArteTVBaseIE._ARTE_LANGUAGES}
+    _TESTS = [{
+        'url': 'https://www.arte.tv/en/videos/088501-000-A/mexico-stealing-petrol-to-survive/',
+        'info_dict': {
+            'id': '088501-000-A',
+            'ext': 'mp4',
+            'title': 'Mexico: Stealing Petrol to Survive',
+            'upload_date': '20190628',
+        },
+    }, {
+        'url': 'https://www.arte.tv/pl/videos/100103-000-A/usa-dyskryminacja-na-porodowce/',
+        'only_matching': True,
+    }, {
+        'url': 'https://api.arte.tv/api/player/v2/config/de/100605-013-A',
+        'only_matching': True,
+    }]
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+        video_id = mobj.group('id')
+        lang = mobj.group('lang') or mobj.group('lang_2')
+
+        info = self._download_json(
+            '%s/config/%s/%s' % (self._API_BASE, lang, video_id), video_id)
         player_info = info['videoJsonPlayer']
 
         vsr = try_get(player_info, lambda x: x['VSR'], dict)
@@ -38,18 +71,11 @@ class ArteTVBaseIE(InfoExtractor):
         if not upload_date_str:
             upload_date_str = (player_info.get('VRA') or player_info.get('VDA') or '').split(' ')[0]
 
-        title = (player_info.get('VTI') or title or player_info['VID']).strip()
+        title = (player_info.get('VTI') or player_info['VID']).strip()
         subtitle = player_info.get('VSU', '').strip()
         if subtitle:
             title += ' - %s' % subtitle
 
-        info_dict = {
-            'id': player_info['VID'],
-            'title': title,
-            'description': player_info.get('VDE'),
-            'upload_date': unified_strdate(upload_date_str),
-            'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
-        }
         qfunc = qualities(['MQ', 'HQ', 'EQ', 'SQ'])
 
         LANGS = {
@@ -64,7 +90,6 @@ class ArteTVBaseIE(InfoExtractor):
         langcode = LANGS.get(lang, lang)
 
         formats = []
-        m3u8_formats = []
         for format_id, format_dict in vsr.items():
             f = dict(format_dict)
             format_url = url_or_none(f.get('url'))
@@ -120,6 +145,7 @@ class ArteTVBaseIE(InfoExtractor):
                     m3u8_id=format_id, fatal=False)
                 for m3u8_format in m3u8_formats:
                     m3u8_format['language_preference'] = lang_pref
+                formats.extend(m3u8_formats)
                 continue
 
             format = {
@@ -142,58 +168,50 @@ class ArteTVBaseIE(InfoExtractor):
 
             formats.append(format)
 
-        self._check_formats(formats, video_id)
-
-        formats.extend(m3u8_formats)
         self._sort_formats(formats)
 
-        info_dict['formats'] = formats
-        return info_dict
-
+        return {
+            'id': player_info.get('VID') or video_id,
+            'title': title,
+            'description': player_info.get('VDE'),
+            'upload_date': unified_strdate(upload_date_str),
+            'thumbnail': player_info.get('programImage') or player_info.get('VTU', {}).get('IUR'),
+            'formats': formats,
+        }
 
-class ArteTVPlus7IE(ArteTVBaseIE):
-    IE_NAME = 'arte.tv:+7'
-    _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?P<lang>fr|de|en|es|it|pl)/videos/(?P<id>\d{6}-\d{3}-[AF])'
 
+class ArteTVEmbedIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?arte\.tv/player/v\d+/index\.php\?.*?\bjson_url=.+'
     _TESTS = [{
-        'url': 'https://www.arte.tv/en/videos/088501-000-A/mexico-stealing-petrol-to-survive/',
+        'url': 'https://www.arte.tv/player/v5/index.php?json_url=https%3A%2F%2Fapi.arte.tv%2Fapi%2Fplayer%2Fv2%2Fconfig%2Fde%2F100605-013-A&lang=de&autoplay=true&mute=0100605-013-A',
         'info_dict': {
-            'id': '088501-000-A',
+            'id': '100605-013-A',
             'ext': 'mp4',
-            'title': 'Mexico: Stealing Petrol to Survive',
-            'upload_date': '20190628',
+            'title': 'United we Stream November Lockdown Edition #13',
+            'description': 'md5:be40b667f45189632b78c1425c7c2ce1',
+            'upload_date': '20201116',
         },
+    }, {
+        'url': 'https://www.arte.tv/player/v3/index.php?json_url=https://api.arte.tv/api/player/v2/config/de/100605-013-A',
+        'only_matching': True,
     }]
 
-    def _real_extract(self, url):
-        lang, video_id = re.match(self._VALID_URL, url).groups()
-        return self._extract_from_json_url(
-            'https://api.arte.tv/api/player/v1/config/%s/%s' % (lang, video_id),
-            video_id, lang)
-
-
-class ArteTVEmbedIE(ArteTVPlus7IE):
-    IE_NAME = 'arte.tv:embed'
-    _VALID_URL = r'''(?x)
-        https://www\.arte\.tv
-        /player/v3/index\.php\?json_url=
-        (?P<json_url>
-            https?://api\.arte\.tv/api/player/v1/config/
-            (?P<lang>[^/]+)/(?P<id>\d{6}-\d{3}-[AF])
-        )
-    '''
-
-    _TESTS = []
+    @staticmethod
+    def _extract_urls(webpage):
+        return [url for _, url in re.findall(
+            r'<(?:iframe|script)[^>]+src=(["\'])(?P<url>(?:https?:)?//(?:www\.)?arte\.tv/player/v\d+/index\.php\?.*?\bjson_url=.+?)\1',
+            webpage)]
 
     def _real_extract(self, url):
-        json_url, lang, video_id = re.match(self._VALID_URL, url).groups()
-        return self._extract_from_json_url(json_url, video_id, lang)
+        qs = compat_urlparse.parse_qs(compat_urlparse.urlparse(url).query)
+        json_url = qs['json_url'][0]
+        video_id = ArteTVIE._match_id(json_url)
+        return self.url_result(
+            json_url, ie=ArteTVIE.ie_key(), video_id=video_id)
 
 
 class ArteTVPlaylistIE(ArteTVBaseIE):
-    IE_NAME = 'arte.tv:playlist'
-    _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?P<lang>fr|de|en|es|it|pl)/videos/(?P<id>RC-\d{6})'
-
+    _VALID_URL = r'https?://(?:www\.)?arte\.tv/(?P<lang>%s)/videos/(?P<id>RC-\d{6})' % ArteTVBaseIE._ARTE_LANGUAGES
     _TESTS = [{
         'url': 'https://www.arte.tv/en/videos/RC-016954/earn-a-living/',
         'info_dict': {
@@ -202,17 +220,35 @@ class ArteTVPlaylistIE(ArteTVBaseIE):
             'description': 'md5:d322c55011514b3a7241f7fb80d494c2',
         },
         'playlist_mincount': 6,
+    }, {
+        'url': 'https://www.arte.tv/pl/videos/RC-014123/arte-reportage/',
+        'only_matching': True,
     }]
 
     def _real_extract(self, url):
         lang, playlist_id = re.match(self._VALID_URL, url).groups()
         collection = self._download_json(
-            'https://api.arte.tv/api/player/v1/collectionData/%s/%s?source=videos'
-            % (lang, playlist_id), playlist_id)
+            '%s/collectionData/%s/%s?source=videos'
+            % (self._API_BASE, lang, playlist_id), playlist_id)
+        entries = []
+        for video in collection['videos']:
+            if not isinstance(video, dict):
+                continue
+            video_url = url_or_none(video.get('url')) or url_or_none(video.get('jsonUrl'))
+            if not video_url:
+                continue
+            video_id = video.get('programId')
+            entries.append({
+                '_type': 'url_transparent',
+                'url': video_url,
+                'id': video_id,
+                'title': video.get('title'),
+                'alt_title': video.get('subtitle'),
+                'thumbnail': url_or_none(try_get(video, lambda x: x['mainImage']['url'], compat_str)),
+                'duration': int_or_none(video.get('durationSeconds')),
+                'view_count': int_or_none(video.get('views')),
+                'ie_key': ArteTVIE.ie_key(),
+            })
         title = collection.get('title')
         description = collection.get('shortDescription') or collection.get('teaserText')
-        entries = [
-            self._extract_from_json_url(
-                video['jsonUrl'], video.get('programId') or playlist_id, lang)
-            for video in collection['videos'] if video.get('jsonUrl')]
         return self.playlist_result(entries, playlist_id, title, description)
index 11ef4726150a43b719e000af039829ec161635f1..088800eb9f404e38e50025640bbb39350f85095c 100644 (file)
@@ -58,7 +58,7 @@ from .ard import (
     ARDMediathekIE,
 )
 from .arte import (
-    ArteTVPlus7IE,
+    ArteTVIE,
     ArteTVEmbedIE,
     ArteTVPlaylistIE,
 )
index 355067a509fc197c10a2085bad058d4459729bc9..d08a8cca58cb36a6af62b43af1fde785a0d5f042 100644 (file)
@@ -91,6 +91,7 @@ from .piksel import PikselIE
 from .videa import VideaIE
 from .twentymin import TwentyMinutenIE
 from .ustream import UstreamIE
+from .arte import ArteTVEmbedIE
 from .videopress import VideoPressIE
 from .rutube import RutubeIE
 from .limelight import LimelightBaseIE
@@ -2760,11 +2761,9 @@ class GenericIE(InfoExtractor):
             return self.url_result(ustream_url, UstreamIE.ie_key())
 
         # Look for embedded arte.tv player
-        mobj = re.search(
-            r'<(?:script|iframe) [^>]*?src="(?P<url>http://www\.arte\.tv/(?:playerv2/embed|arte_vp/index)[^"]+)"',
-            webpage)
-        if mobj is not None:
-            return self.url_result(mobj.group('url'), 'ArteTVEmbed')
+        arte_urls = ArteTVEmbedIE._extract_urls(webpage)
+        if arte_urls:
+            return self.playlist_from_matches(arte_urls, video_id, video_title)
 
         # Look for embedded francetv player
         mobj = re.search(