Merge 1058c29c74 into c5098961b0

Now also succeeds with player b12cc44b

test/test_jsinterp.py

@@ -425,6 +425,34 @@ class TestJSInterpreter(unittest.TestCase):
             self._test(jsi, [''], args=['', '-'])
             self._test(jsi, [], args=['', ''])
 
+    def test_slice(self):
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice()}', [0, 1, 2, 3, 4, 5, 6, 7, 8])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(0)}', [0, 1, 2, 3, 4, 5, 6, 7, 8])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(5)}', [5, 6, 7, 8])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(99)}', [])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-2)}', [7, 8])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-99)}', [0, 1, 2, 3, 4, 5, 6, 7, 8])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(0, 0)}', [])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(1, 0)}', [])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(0, 1)}', [0])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(3, 6)}', [3, 4, 5])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(1, -1)}', [1, 2, 3, 4, 5, 6, 7])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-1, 1)}', [])
+        self._test('function f(){return [0, 1, 2, 3, 4, 5, 6, 7, 8].slice(-3, -1)}', [6, 7])
+        self._test('function f(){return "012345678".slice()}', '012345678')
+        self._test('function f(){return "012345678".slice(0)}', '012345678')
+        self._test('function f(){return "012345678".slice(5)}', '5678')
+        self._test('function f(){return "012345678".slice(99)}', '')
+        self._test('function f(){return "012345678".slice(-2)}', '78')
+        self._test('function f(){return "012345678".slice(-99)}', '012345678')
+        self._test('function f(){return "012345678".slice(0, 0)}', '')
+        self._test('function f(){return "012345678".slice(1, 0)}', '')
+        self._test('function f(){return "012345678".slice(0, 1)}', '0')
+        self._test('function f(){return "012345678".slice(3, 6)}', '345')
+        self._test('function f(){return "012345678".slice(1, -1)}', '1234567')
+        self._test('function f(){return "012345678".slice(-1, 1)}', '')
+        self._test('function f(){return "012345678".slice(-3, -1)}', '67')
+
 
 if __name__ == '__main__':
     unittest.main()

test/test_youtube_signature.py

@@ -174,6 +174,14 @@ _NSIG_TESTS = [
         'https://www.youtube.com/s/player/5604538d/player_ias.vflset/en_US/base.js',
         '7X-he4jjvMx7BCX', 'sViSydX8IHtdWA',
     ),
+    (
+        'https://www.youtube.com/s/player/20dfca59/player_ias.vflset/en_US/base.js',
+        '-fLCxedkAk4LUTK2', 'O8kfRq1y1eyHGw',
+    ),
+    (
+        'https://www.youtube.com/s/player/b12cc44b/player_ias.vflset/en_US/base.js',
+        'keLa5R2U00sR9SQK', 'N1OGyujjEwMnLw',
+    ),
 ]
 
 

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',
+        'url': 'https://www.rtp.pt/play/p117/e563265/os-contemporaneos',
-        'md5': 'e736ce0c665e459ddb818546220b4ef8',
         'info_dict': {
-            'id': 'e174042',
+            'id': 'os-contemporaneos',
-            'ext': 'mp3',
+            'ext': 'mp4',
-            'title': 'Paixões Cruzadas',
+            'title': 'Os Contemporâneos Episódio 1',
-            'description': 'As paixões musicais de António Cartaxo e António Macedo',
+            'description': 'Os Contemporâneos, um programa de humor com um olhar na sociedade portuguesa!',
-            'thumbnail': r're:^https?://.*\.jpg',
+            '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(
+        # Remove JS multi-line comments from webpage source
-            r'(?s)RTPPlayer\(({.+?})\);', webpage,
+        webpage = re.sub(r'(\/\*.*\*\/)', '', webpage, flags=re.DOTALL)
-            'player config'), video_id, js_to_json)
+
+        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)
+                m3u8_id='hls')
-            if file_key:
+        elif ext == 'm3u8':
-                formats.append({
+            # It can be downloaded without any further changes
-                    'url': 'https://cdn-ondemand.rtp.pt' + file_key,
+            formats = self._extract_m3u8_formats(
-                    'preference': 1,
+                file_url, video_id, 'mp4', 'm3u8_native',
-                })
+                m3u8_id='hls')
-            self._sort_formats(formats)
         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),
+            'subtitles': subtitles,
-            'thumbnail': config.get('poster') or self._og_search_thumbnail(webpage),
+            '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

@@ -1659,17 +1659,46 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
     def _extract_n_function_name(self, jscode):
         func_name, idx = self._search_regex(
             # new: (b=String.fromCharCode(110),c=a.get(b))&&c=nfunc[idx](c)
-            # or:  (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c)s
+            # or:  (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c)
-            # old: .get("n"))&&(b=nfunc[idx](b)
+            # or:  (PL(a),b=a.j.n||null)&&(b=nfunc[idx](b)
-            # older: .get("n"))&&(b=nfunc(b)
+            # or:  (b="nn"[+a.D],vL(a),c=a.j[b]||null)&&(c=narray[idx](c),a.set(b,c),narray.length||nfunc("")
+            # old: (b=a.get("n"))&&(b=nfunc[idx](b)(?P<c>[a-z])\s*=\s*[a-z]\s*
+            # older: (b=a.get("n"))&&(b=nfunc(b)
             r'''(?x)
-                (?:\(\s*(?P<b>[a-z])\s*=\s*(?:
+                \((?:[\w$()\s]+,)*?\s*      # (
+                (?P<b>[a-z])\s*=\s*         # b=
+                (?:
+                    (?:                     # expect ,c=a.get(b) (etc)
                         String\s*\.\s*fromCharCode\s*\(\s*110\s*\)|
                         "n+"\[\s*\+?s*[\w$.]+\s*]
-                )\s*,(?P<c>[a-z])\s*=\s*[a-z]\s*)?
+                    )\s*(?:,[\w$()\s]+(?=,))*|
-                \.\s*get\s*\(\s*(?(b)(?P=b)|"n{1,2}")(?:\s*\)){2}\s*&&\s*\(\s*(?(c)(?P=c)|b)\s*=\s*
+                       (?P<old>[\w$]+)      # a (old[er])
+                   )\s*
+                   (?(old)
+                                            # b.get("n")
+                       (?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*?
+                       (?:\.\s*n|\[\s*"n"\s*]|\.\s*get\s*\(\s*"n"\s*\))
+                       |                    # ,c=a.get(b)
+                       ,\s*(?P<c>[a-z])\s*=\s*[a-z]\s*
+                       (?:\.\s*[\w$]+\s*|\[\s*[\w$]+\s*]\s*)*?
+                       (?:\[\s*(?P=b)\s*]|\.\s*get\s*\(\s*(?P=b)\s*\))
+                   )
+                                            # interstitial junk
+                   \s*(?:\|\|\s*null\s*)?(?:\)\s*)?&&\s*(?:\(\s*)?
+               (?(c)(?P=c)|(?P=b))\s*=\s*   # [c|b]=
+                                            # nfunc|nfunc[idx]
                    (?P<nfunc>[a-zA-Z_$][\w$]*)(?:\s*\[(?P<idx>\d+)\])?\s*\(\s*[\w$]+\s*\)
-            ''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'))
+            ''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'),
+            default=(None, None))
+        # thx bashonly: yt-dlp/yt-dlp/pull/10611
+        if not func_name:
+            self.report_warning('Falling back to generic n function search')
+            return self._search_regex(
+                r'''(?xs)
+                    (?:(?<=[^\w$])|^)       # instead of \b, which ignores $
+                    (?P<name>(?!\d)[a-zA-Z\d_$]+)\s*=\s*function\((?!\d)[a-zA-Z\d_$]+\)
+                    \s*\{(?:(?!};).)+?["']enhanced_except_
+                ''', jscode, 'Initial JS player n function name', group='name')
         if not idx:
             return func_name
 

youtube_dl/jsinterp.py

@@ -925,9 +925,16 @@ class JSInterpreter(object):
                     obj.reverse()
                     return obj
                 elif member == 'slice':
-                    assertion(isinstance(obj, list), 'must be applied on a list')
+                    assertion(isinstance(obj, (list, compat_str)), 'must be applied on a list or string')
-                    assertion(len(argvals) == 1, 'takes exactly one argument')
+                    # From [1]:
-                    return obj[argvals[0]:]
+                    # .slice() - like [:]
+                    # .slice(n) - like [n:] (not [slice(n)]
+                    # .slice(m, n) - like [m:n] or [slice(m, n)]
+                    # [1] https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Array/slice
+                    assertion(len(argvals) <= 2, 'takes between 0 and 2 arguments')
+                    if len(argvals) < 2:
+                        argvals += (None,)
+                    return obj[slice(*argvals)]
                 elif member == 'splice':
                     assertion(isinstance(obj, list), 'must be applied on a list')
                     assertion(argvals, 'takes one or more arguments')