Merge 1058c29c74 into 1036478d13

* WEB URLs are, MWEB not
* resolves #33017

youtube_dl/extractor/rtp.py

@@ -3,25 +3,71 @@ from __future__ import unicode_literals
 
 from .common import InfoExtractor
 from ..utils import (
+    ExtractorError,
     determine_ext,
     js_to_json,
 )
+from ..compat import (
+    compat_b64decode,
+    compat_urllib_parse_unquote,
+)
+
+import re
 
 
 class RTPIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?rtp\.pt/play/p(?P<program_id>[0-9]+)/(?P<id>[^/?#]+)/?'
+    _VALID_URL = r'https?://(?:(?:(?:www\.)?rtp\.pt/play/(?P<subarea>.*/)?p(?P<program_id>[0-9]+)/(?P<episode_id>e[0-9]+/)?)|(?:arquivos\.rtp\.pt/conteudos/))(?P<id>[^/?#]+)/?'
     _TESTS = [{
-        'url': 'http://www.rtp.pt/play/p405/e174042/paixoes-cruzadas',
-        'md5': 'e736ce0c665e459ddb818546220b4ef8',
+        'url': 'https://www.rtp.pt/play/p117/e563265/os-contemporaneos',
         'info_dict': {
-            'id': 'e174042',
-            'ext': 'mp3',
-            'title': 'Paixões Cruzadas',
-            'description': 'As paixões musicais de António Cartaxo e António Macedo',
-            'thumbnail': r're:^https?://.*\.jpg',
+            'id': 'os-contemporaneos',
+            'ext': 'mp4',
+            'title': 'Os Contemporâneos Episódio 1',
+            'description': 'Os Contemporâneos, um programa de humor com um olhar na sociedade portuguesa!',
+            'thumbnail': r're:^https?://.*\.(jpg|png)'
         },
     }, {
-        'url': 'http://www.rtp.pt/play/p831/a-quimica-das-coisas',
+        'url': 'https://www.rtp.pt/play/p8157/e541212/telejornal',
+        'info_dict': {
+            'id': 'telejornal',
+            'ext': 'mp4',
+            'title': 'Telejornal de 01 Mai 2021 PARTE 1',
+            'description': 'A mais rigorosa seleção de notícias, todos os dias às 20h00. De segunda a domingo, João Adelino Faria e José Rodrigues dos Santos mostram-lhe o que de'
+        },
+    }, {
+        'url': 'https://www.rtp.pt/play/p6646/e457262/grande-entrevista',
+        'info_dict': {
+            'id': 'grande-entrevista',
+            'ext': 'mp4',
+            'title': 'Grande Entrevista Episódio 7 - de 19 Fev 2020',
+            'description': 'Bruno Nogueira - É um dos mais originais humoristas portugueses e de maior êxito! Bruno Nogueira na Grande Entrevista com Vítor Gonçalves.'
+        },
+    }, {
+        'url': 'https://www.rtp.pt/play/estudoemcasa/p7776/e539826/portugues-1-ano',
+        'info_dict': {
+            'id': 'portugues-1-ano',
+            'ext': 'mp4',
+            'title': 'Português - 1.º ano , aula 45 -  27 Abr 2021 - Estudo Em Casa - RTP',
+            'description': 'A História do Pedrito Coelho, de Beatrix Potter. O dígrafo \'lh\' - A História do Pedrito Coelho, de Beatrix Potter. O dígrafo \'lh\'.'
+        },
+    }, {
+        'url': 'https://www.rtp.pt/play/zigzag/p5449/e385973/banda-zig-zag',
+        'info_dict': {
+            'id': 'banda-zig-zag',
+            'ext': 'mp4',
+            'title': 'Banda Zig Zag Episódio 1 -  Zig Zag Play - RTP',
+            'description': 'A Amizade é o Nosso Mel - Zig: é a menina que além de tocar também canta. Adora aprender palavras novas e adora ler. Gosta de fazer palavras cruzadas'
+        },
+    }, {
+        'url': 'https://arquivos.rtp.pt/conteudos/liga-dos-ultimos-152/',
+        'info_dict': {
+            'id': 'liga-dos-ultimos-152',
+            'ext': 'mp4',
+            'title': 'Liga dos Últimos – RTP Arquivos',
+            'description': 'Magazine desportivo, com apresentação de Álvaro Costa e comentários em estúdio do professor Hernâni Gonçalves e do sociólogo João Nuno Coelho. Destaque para os jogos de futebol das equipas dos escalões secundários de Portugal, com momentos dos jogos: Agrário de Lamas vs Pampilhoense e Apúlia vs Fragoso.'
+        },
+    }, {
+        'url': 'https://www.rtp.pt/play/p510/aleixo-fm',
         'only_matching': True,
     }]
 
@@ -29,38 +75,115 @@ class RTPIE(InfoExtractor):
         video_id = self._match_id(url)
 
         webpage = self._download_webpage(url, video_id)
-        title = self._html_search_meta(
-            'twitter:title', webpage, display_name='title', fatal=True)
 
-        config = self._parse_json(self._search_regex(
-            r'(?s)RTPPlayer\(({.+?})\);', webpage,
-            'player config'), video_id, js_to_json)
+        # Remove JS multi-line comments from webpage source
+        webpage = re.sub(r'(\/\*.*\*\/)', '', webpage, flags=re.DOTALL)
+
+        title = self._html_search_regex(r'<title>(.+?)</title>', webpage, 'title')
+
+        if 'Este episódio não se encontra disponível' in title:
+            raise ExtractorError('Episode unavailable', expected=True)
+
+        # Replace irrelevant text in title
+        title = re.sub(r' -  ?RTP Play - RTP', '', title)
+
+        # Check if it's a video split in parts, if so add part number to title
+        part = self._html_search_regex(r'section\-parts.*<span.*>(.+?)</span>.*</ul>', webpage, 'part', default=None)
+        if part:
+            title = '{title} {part}'.format(title=title, part=part)
+
+        # Get JS object
+        js_object = self._search_regex(r'(?s)RTPPlayer *\( *({.+?}) *\);', webpage, 'player config')
+
+        json_string_for_config = ''
+        filekey_found = False
+
+        # Verify JS object since it isn't pure JSON and probably needs some fixing
+        for line in js_object.splitlines():
+            stripped_line = line.strip()
+
+            # If JS object key is 'fileKey'
+            if re.match('fileKey ?:', stripped_line):
+                filekey_found = True
+                if 'decodeURIComponent' in stripped_line:
+                    # 1) The value is an encoded URL
+                    encoded_url = re.match(r"[^[]*\[([^]]*)\]", stripped_line).groups()[0]
+                    encoded_url = re.sub(r'[\s"\',]', '', encoded_url)
+
+                    if 'atob' in stripped_line:
+                        # Most of the times 'atob' approach is used but not always so we need to be sure
+                        decoded_url = compat_b64decode(
+                            compat_urllib_parse_unquote(
+                                encoded_url)).decode('utf-8')
+                    else:
+                        # If no 'atob' we just need to unquote it
+                        decoded_url = compat_urllib_parse_unquote(encoded_url)
+
+                    # Insert the (relative) decoded URL in JSON
+                    json_string_for_config += '\nfileKey: "{decoded_url}",'.format(decoded_url=decoded_url)
+                else:
+                    # 2) ... or the value URL is not encoded so keep it that way
+                    json_string_for_config += '\n{stripped_line}'.format(stripped_line=stripped_line)
+
+            elif (
+                not stripped_line.startswith("//")
+                and not re.match('.*extraSettings ?:', stripped_line)
+                and (not filekey_found or (filekey_found and not re.match('file ?:', stripped_line)))
+            ):
+                # Ignore commented lines and 'extraSettings'. Also ignore 'file' if 'fileKey' already exists
+                json_string_for_config += '\n{stripped_line}'.format(stripped_line=stripped_line)
+
+        # Finally send pure JSON string for JSON parsing
+        config = self._parse_json(json_string_for_config, video_id, js_to_json)
+
+        if 'fileKey' in config:
+            # 'fileKey' has priority over 'file' on our end
+            file_url = config['fileKey']
+        elif 'file' in config:
+            # 'RTP Arquivos' still uses old regular non-encoded 'file' key
             file_url = config['file']
+        else:
+            raise ExtractorError('No valid media source found in page')
+
         ext = determine_ext(file_url)
-        if ext == 'm3u8':
-            file_key = config.get('fileKey')
+
+        if ext == 'mp4':
+            # Due to recent changes, we need to hardcode the URL like this and download it using 'm3u8'
+            file_url = 'https://streaming-vod.rtp.pt/hls{file_url}/index-v1-a1.m3u8'.format(file_url=file_url)
+
             formats = self._extract_m3u8_formats(
                 file_url, video_id, 'mp4', 'm3u8_native',
-                m3u8_id='hls', fatal=file_key)
-            if file_key:
-                formats.append({
-                    'url': 'https://cdn-ondemand.rtp.pt' + file_key,
-                    'preference': 1,
-                })
-            self._sort_formats(formats)
+                m3u8_id='hls')
+        elif ext == 'm3u8':
+            # It can be downloaded without any further changes
+            formats = self._extract_m3u8_formats(
+                file_url, video_id, 'mp4', 'm3u8_native',
+                m3u8_id='hls')
         else:
+            # Need to set basepath
+            file_url = 'https://cdn-ondemand.rtp.pt{file_url}'.format(file_url=file_url)
             formats = [{
                 'url': file_url,
                 'ext': ext,
             }]
-        if config.get('mediaType') == 'audio':
+
+        if config['mediaType'] == 'audio':
             for f in formats:
                 f['vcodec'] = 'none'
 
+        subtitles = {}
+        if 'vtt' in config:
+            sub_lang, sub_lang_full, sub_url = config['vtt'][0]
+            subtitles.setdefault(sub_lang, []).append({
+                'url': sub_url,
+                'ext': 'vtt',
+            })
+
         return {
             'id': video_id,
             'title': title,
             'formats': formats,
-            'description': self._html_search_meta(['description', 'twitter:description'], webpage),
-            'thumbnail': config.get('poster') or self._og_search_thumbnail(webpage),
+            'subtitles': subtitles,
+            'description': self._html_search_meta(['og:description', 'description', 'twitter:description'], webpage),
+            'thumbnail': config['poster'] or self._og_search_thumbnail(webpage),
         }

youtube_dl/extractor/youtube.py

@@ -9,6 +9,7 @@ import json
 import os.path
 import random
 import re
+import string
 import time
 import traceback
 
@@ -67,6 +68,7 @@ from ..utils import (
 
 class YoutubeBaseInfoExtractor(InfoExtractor):
     """Provide base functions for Youtube extractors"""
+
     _LOGIN_URL = 'https://accounts.google.com/ServiceLogin'
     _TWOFACTOR_URL = 'https://accounts.google.com/signin/challenge'
 
@@ -138,7 +140,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
                 [2, 1, None, 1,
                  'https://accounts.google.com/ServiceLogin?passive=true&continue=https%3A%2F%2Fwww.youtube.com%2Fsignin%3Fnext%3D%252F%26action_handle_signin%3Dtrue%26hl%3Den%26app%3Ddesktop%26feature%3Dsign_in_button&hl=en&service=youtube&uilel=3&requestPath=%2FServiceLogin&Page=PasswordSeparationSignIn',
                  None, [], 4],
-                1, [None, None, []], None, None, None, True
+                1, [None, None, []], None, None, None, True,
             ],
             username,
         ]
@@ -160,7 +162,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             None, 1, None, [1, None, None, None, [password, None, True]],
             [
                 None, None, [2, 1, None, 1, 'https://accounts.google.com/ServiceLogin?passive=true&continue=https%3A%2F%2Fwww.youtube.com%2Fsignin%3Fnext%3D%252F%26action_handle_signin%3Dtrue%26hl%3Den%26app%3Ddesktop%26feature%3Dsign_in_button&hl=en&service=youtube&uilel=3&requestPath=%2FServiceLogin&Page=PasswordSeparationSignIn', None, [], 4],
-                1, [None, None, []], None, None, None, True
+                1, [None, None, []], None, None, None, True,
             ]]
 
         challenge_results = req(
@@ -213,7 +215,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
                     user_hash, None, 2, None,
                     [
                         9, None, None, None, None, None, None, None,
-                        [None, tfa_code, True, 2]
+                        [None, tfa_code, True, 2],
                     ]]
 
                 tfa_results = req(
@@ -284,7 +286,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             'client': {
                 'clientName': 'WEB',
                 'clientVersion': '2.20201021.03.00',
-            }
+            },
         },
     }
 
@@ -385,7 +387,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
                 'client': {
                     'clientName': 'WEB',
                     'clientVersion': '2.20201021.03.00',
-                }
+                },
             },
             'query': query,
         }
@@ -462,7 +464,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor):
             #       (HTML, videodetails, metadata, renderers)
             'name': ('content', 'author', (('ownerChannelName', None), 'title'), ['text']),
             'url': ('href', 'ownerProfileUrl', 'vanityChannelUrl',
-                    ['navigationEndpoint', 'browseEndpoint', 'canonicalBaseUrl'])
+                    ['navigationEndpoint', 'browseEndpoint', 'canonicalBaseUrl']),
         }
         if any((videodetails, metadata, renderers)):
             result = (
@@ -671,7 +673,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/user/8KVIDEO',
                 'description': '',
                 'uploader': '8KVIDEO',
-                'title': 'UHDTV TEST 8K VIDEO.mp4'
+                'title': 'UHDTV TEST 8K VIDEO.mp4',
             },
             'params': {
                 'youtube_include_dash_manifest': True,
@@ -711,7 +713,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'uploader_url': r're:https?://(?:www\.)?youtube\.com/@theamazingatheist',
                 'title': 'Burning Everyone\'s Koran',
                 'description': 'SUBSCRIBE: http://www.youtube.com/saturninefilms \r\n\r\nEven Obama has taken a stand against freedom on this issue: http://www.huffingtonpost.com/2010/09/09/obama-gma-interview-quran_n_710282.html',
-            }
+            },
         },
         # Age-gated videos
         {
@@ -839,7 +841,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             },
             'expected_warnings': [
                 'DASH manifest missing',
-            ]
+            ],
         },
         # Olympics (https://github.com/ytdl-org/youtube-dl/issues/4431)
         {
@@ -1820,8 +1822,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
 
         # cpn generation algorithm is reverse engineered from base.js.
         # In fact it works even with dummy cpn.
-        CPN_ALPHABET = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_'
-        cpn = ''.join(CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(0, 16))
+        CPN_ALPHABET = string.ascii_letters + string.digits + '-_'
+        cpn = ''.join(CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(16))
 
         # more consistent results setting it to right before the end
         qs = parse_qs(playback_url)
@@ -1881,8 +1883,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         mobj = re.match(cls._VALID_URL, url, re.VERBOSE)
         if mobj is None:
             raise ExtractorError('Invalid URL: %s' % url)
-        video_id = mobj.group(2)
-        return video_id
+        return mobj.group(2)
 
     def _extract_chapters_from_json(self, data, video_id, duration):
         chapters_list = try_get(
@@ -2035,7 +2036,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             headers = {
                 'X-YouTube-Client-Name': '85',
                 'X-YouTube-Client-Version': '2.0',
-                'Origin': 'https://www.youtube.com'
+                'Origin': 'https://www.youtube.com',
             }
 
             video_info = self._call_api('player', query, video_id, fatal=False, headers=headers)
@@ -2064,8 +2065,8 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             return ''.join([r['text'] for r in runs if isinstance(r.get('text'), compat_str)])
 
         search_meta = (
-            lambda x: self._html_search_meta(x, webpage, default=None)) \
-            if webpage else lambda x: None
+            (lambda x: self._html_search_meta(x, webpage, default=None))
+            if webpage else lambda _: None)
 
         video_details = player_response.get('videoDetails') or {}
         microformat = try_get(
@@ -2137,7 +2138,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         def build_fragments(f):
             return LazyList({
                 'url': update_url_query(f['url'], {
-                    'range': '{0}-{1}'.format(range_start, min(range_start + CHUNK_SIZE - 1, f['filesize']))
+                    'range': '{0}-{1}'.format(range_start, min(range_start + CHUNK_SIZE - 1, f['filesize'])),
                 })
             } for range_start in range(0, f['filesize'], CHUNK_SIZE))
 
@@ -2236,7 +2237,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                     'protocol': 'http_dash_segments',
                     'fragments': build_fragments(dct),
                 } if dct['filesize'] else {
-                    'downloader_options': {'http_chunk_size': CHUNK_SIZE}  # No longer useful?
+                    'downloader_options': {'http_chunk_size': CHUNK_SIZE},  # No longer useful?
                 })
 
             formats.append(dct)
@@ -2414,9 +2415,9 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             'is_live': is_live,
         }
 
-        pctr = try_get(
+        pctr = traverse_obj(
             player_response,
-            lambda x: x['captions']['playerCaptionsTracklistRenderer'], dict)
+            ('captions', 'playerCaptionsTracklistRenderer', T(dict)))
         if pctr:
             def process_language(container, base_url, lang_code, query):
                 lang_subs = []
@@ -2430,9 +2431,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                     })
                 container[lang_code] = lang_subs
 
+            def process_subtitles():
                 subtitles = {}
-            for caption_track in (pctr.get('captionTracks') or []):
-                base_url = caption_track.get('baseUrl')
+                for caption_track in traverse_obj(pctr, (
+                        'captionTracks', lambda _, v: v.get('baseUrl'))):
+                    base_url = self._yt_urljoin(caption_track['baseUrl'])
                     if not base_url:
                         continue
                     if caption_track.get('kind') != 'asr':
@@ -2443,18 +2446,19 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                             subtitles, base_url, lang_code, {})
                         continue
                     automatic_captions = {}
-                for translation_language in (pctr.get('translationLanguages') or []):
-                    translation_language_code = translation_language.get('languageCode')
-                    if not translation_language_code:
-                        continue
+                    for translation_language in traverse_obj(pctr, (
+                            'translationLanguages', lambda _, v: v.get('languageCode'))):
+                        translation_language_code = translation_language['languageCode']
                         process_language(
                             automatic_captions, base_url, translation_language_code,
                             {'tlang': translation_language_code})
                     info['automatic_captions'] = automatic_captions
                 info['subtitles'] = subtitles
 
+            process_subtitles()
+
         parsed_url = compat_urllib_parse_urlparse(url)
-        for component in [parsed_url.fragment, parsed_url.query]:
+        for component in (parsed_url.fragment, parsed_url.query):
             query = compat_parse_qs(component)
             for k, v in query.items():
                 for d_k, s_ks in [('start', ('start', 't')), ('end', ('end',))]:
@@ -2684,7 +2688,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
             'title': 'Super Cooper Shorts - Shorts',
             'uploader': 'Super Cooper Shorts',
             'uploader_id': '@SuperCooperShorts',
-        }
+        },
     }, {
         # Channel that does not have a Shorts tab. Test should just download videos on Home tab instead
         'url': 'https://www.youtube.com/@emergencyawesome/shorts',
@@ -2738,7 +2742,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
             'description': 'md5:609399d937ea957b0f53cbffb747a14c',
             'uploader': 'ThirstForScience',
             'uploader_id': '@ThirstForScience',
-        }
+        },
     }, {
         'url': 'https://www.youtube.com/c/ChristophLaimer/playlists',
         'only_matching': True,
@@ -3037,7 +3041,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
             'uploader': '3Blue1Brown',
             'uploader_id': '@3blue1brown',
             'channel_id': 'UCYO_jab_esuFRV4b17AJtAw',
-        }
+        },
     }]
 
     @classmethod
@@ -3335,7 +3339,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
             'client': {
                 'clientName': 'WEB',
                 'clientVersion': client_version,
-            }
+            },
         }
         visitor_data = try_get(context, lambda x: x['client']['visitorData'], compat_str)
 
@@ -3354,7 +3358,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
                 headers['x-goog-visitor-id'] = visitor_data
             data['continuation'] = continuation['continuation']
             data['clickTracking'] = {
-                'clickTrackingParams': continuation['itct']
+                'clickTrackingParams': continuation['itct'],
             }
             count = 0
             retries = 3
@@ -3613,7 +3617,7 @@ class YoutubePlaylistIE(InfoExtractor):
             'uploader': 'milan',
             'uploader_id': '@milan5503',
             'channel_id': 'UCEI1-PVPcYXjB73Hfelbmaw',
-        }
+        },
     }, {
         'url': 'http://www.youtube.com/embed/_xDOZElKyNU?list=PLsyOSbh5bs16vubvKePAQ1x3PhKavfBIl',
         'playlist_mincount': 455,
@@ -3623,7 +3627,7 @@ class YoutubePlaylistIE(InfoExtractor):
             'uploader': 'LBK',
             'uploader_id': '@music_king',
             'channel_id': 'UC21nz3_MesPLqtDqwdvnoxA',
-        }
+        },
     }, {
         'url': 'TLGGrESM50VT6acwMjAyMjAxNw',
         'only_matching': True,
@@ -3734,7 +3738,7 @@ class YoutubeSearchIE(SearchInfoExtractor, YoutubeBaseInfoExtractor):
         'info_dict': {
             'id': 'youtube-dl test video',
             'title': 'youtube-dl test video',
-        }
+        },
     }]
 
     def _get_n_results(self, query, n):
@@ -3754,7 +3758,7 @@ class YoutubeSearchDateIE(YoutubeSearchIE):
         'info_dict': {
             'id': 'youtube-dl test video',
             'title': 'youtube-dl test video',
-        }
+        },
     }]
 
 
@@ -3769,7 +3773,7 @@ class YoutubeSearchURLIE(YoutubeBaseInfoExtractor):
             'id': 'youtube-dl test video',
             'title': 'youtube-dl test video',
         },
-        'params': {'playlistend': 5}
+        'params': {'playlistend': 5},
     }, {
         'url': 'https://www.youtube.com/results?q=test&sp=EgQIBBgB',
         'only_matching': True,
@@ -3785,6 +3789,7 @@ class YoutubeSearchURLIE(YoutubeBaseInfoExtractor):
 class YoutubeFeedsInfoExtractor(YoutubeTabIE):
     """
     Base class for feed extractors
+
     Subclasses must define the _FEED_NAME property.
     """
     _LOGIN_REQUIRED = True