Merge 47f1d70149 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/extractors.py

@@ -678,6 +678,10 @@ from .mediasite import (
     MediasiteCatalogIE,
     MediasiteNamedCatalogIE,
 )
+from .mediathekviewweb import (
+    MediathekViewWebSearchIE,
+    MediathekViewWebIE,
+)
 from .medici import MediciIE
 from .megaphone import MegaphoneIE
 from .meipai import MeipaiIE

youtube_dl/extractor/mediathekviewweb.py

@@ -0,0 +1,280 @@
+# coding=utf-8
+from __future__ import unicode_literals
+
+import datetime
+import itertools
+import json
+
+from .common import InfoExtractor, SearchInfoExtractor
+from ..compat import compat_parse_qs, compat_urlparse
+from ..utils import ExtractorError, int_or_none
+
+
+class MediathekViewWebSearchIE(SearchInfoExtractor):
+    IE_NAME = 'mediathekviewweb:search'
+    IE_DESC = 'MediathekViewWeb search'
+    _SEARCH_KEY = 'mvwsearch'
+    _MAX_RESULTS = float('inf')
+    _MAX_RESULTS_PER_PAGE = 50
+
+    _TESTS = [
+        {
+            'url': 'mvwsearchall:sandmännchen !kika',
+            'info_dict': {
+                'title': 'Unser Sandmännchen',
+            },
+            'playlist': [],
+            'playlist_count': 7,
+        },
+        {
+            # Audio description & common topic.
+            'url': 'mvwsearch:#Sendung,Maus Audiodeskription',
+            'info_dict': {
+                'title': 'Die Sendung mit der Maus',
+            },
+            'playlist': [],
+            'playlist_count': 1,
+            'params': {
+                'format': 'medium-audio_description',
+                'skip_download': True,
+            }
+        },
+        {
+            # Sign language.
+            'url': 'mvwsearchall:!ard #Tagesschau Gebärdensprache',
+            'info_dict': {
+                'title': '!ard #Tagesschau Gebärdensprache',
+            },
+            'playlist': [],
+            'playlist_mincount': 365,
+            'params': {
+                'format': 'medium-sign_language',
+                'skip_download': True,
+            },
+        },
+    ]
+
+    # Map of title affixes indicating video variants.
+    _variants = {
+        'audio_description': 'Audiodeskription',
+        'sign_language': 'mit Gebärdensprache',
+    }
+    _future = True
+    _everywhere = False
+
+    def _build_conditions(self, search):
+        # @note So far, there is no API endpoint to convert a query string into
+        #   a complete query object, as required by the /api/query endpoint.
+        filters = {}
+        extra = {}
+
+        for component in search.lower().split():
+            if len(component) == 0:
+                continue
+
+            operator = component[0:1]
+            value = component[1:]
+            if len(value) == 0:
+                # Treat single character query as such.
+                # @note This differs from MVW's implementation.
+                operator = ''
+                value = component
+
+            # Extra, non-field settings.
+            if operator == '>':
+                value = int(value.split(',')[0]) * 60
+                extra['duration_min'] = max(extra.get('duration_min', 0), value)
+                continue
+            elif operator == '<':
+                value = int(value.split(',')[0]) * 60
+                extra['duration_max'] = min(extra.get('duration_max', float('inf')), value)
+                continue
+
+            # Field query operators.
+            if operator == '!':
+                field = 'channel'
+            elif operator == '#':
+                field = 'topic'
+            elif operator == '+':
+                field = 'title'
+            elif operator == '*':
+                field = 'description'
+            else:
+                # No known operator specified.
+                field = 'generic'
+                value = component
+
+            # @note In theory, comma-joined values are for AND queries. However
+            #   so far, each condition is AND joined, even without comma.
+            filters.setdefault(field, []).append(' '.join(value.split(',')))
+
+        # Generic filters can apply to different fields, based on the query.
+        if 'generic' in filters:
+            if self._everywhere:
+                filters['channel,topic,title,description'] = filters['generic']
+            elif 'topic' in filters:
+                filters['title'] = filters['generic']
+            else:
+                filters['topic,title'] = filters['generic']
+            filters.pop('generic')
+
+        conditions = []
+        for field, keys in filters.items():
+            for query in keys:
+                conditions.append({
+                    'fields': field.split(','),
+                    'query': query,
+                })
+
+        return conditions, extra
+
+    def _extract_playlist_entries(self, results):
+        entries = []
+        for item in results:
+            variant = None
+            for key, value in self._variants.items():
+                if item.setdefault('title', '').find(value) != -1:
+                    variant = key
+
+            formats = []
+            formats.append({
+                'url': item['url_video'],
+                'format': ('medium (' + self._variants[variant] + ')') if variant else None,
+                'format_id': ('medium-' + variant) if variant else 'medium',
+                'language_preference': -10 if variant else 10,
+                'quality': -2,
+                'filesize': item.get('size'),
+            })
+            if len(item.get('url_video_low', '')) > 0:
+                formats.append({
+                    'url': item['url_video_low'],
+                    'format': ('low (' + self._variants[variant] + ')') if variant else None,
+                    'format_id': ('low-' + variant) if variant else 'low',
+                    'language_preference': -10 if variant else 10,
+                    'quality': -3,
+                })
+            if len(item.get('url_video_hd', '')) > 0:
+                formats.append({
+                    'url': item['url_video_hd'],
+                    'format': ('high (' + self._variants[variant] + ')') if variant else None,
+                    'format_id': ('high-' + variant) if variant else 'high',
+                    'language_preference': -10 if variant else 10,
+                    'quality': -1,
+                })
+            self._sort_formats(formats)
+
+            video = {
+                '_type': 'video',
+                'formats': formats,
+                'id': item.get('id'),
+                'title': item.get('title'),
+                'description': item.get('description'),
+                'series': item.get('topic'),
+                'channel': item.get('channel'),
+                'uploader': item.get('channel'),
+                'duration': int_or_none(item.get('duration')),
+                'webpage_url': item.get('url_website'),
+            }
+
+            if item.get('timestamp'):
+                upload_date = datetime.datetime.utcfromtimestamp(item['timestamp'])
+                video['upload_date'] = upload_date.strftime('%Y%m%d')
+            if item.get('url_subtitle'):
+                video.setdefault('subtitles', {}).setdefault('de', []).append({
+                    'url': item.get('url_subtitle'),
+                    'ext': 'ttml',
+                })
+            entries.append(video)
+
+        return entries
+
+    def _get_n_results(self, query, n):
+        queries, extra = self._build_conditions(query)
+        queryObject = {
+            'queries': queries,
+            'sortBy': 'timestamp',
+            'sortOrder': 'desc',
+            'future': self._future,
+            'duration_min': extra.get('duration_min'),
+            'duration_max': extra.get('duration_max'),
+            'offset': 0,
+            'size': min(n, self._MAX_RESULTS_PER_PAGE),
+        }
+
+        entries = []
+        for page_num in itertools.count(1):
+            queryObject.update({'offset': (page_num - 1) * queryObject['size']})
+            results = self._download_json('https://mediathekviewweb.de/api/query', query,
+                                          note='Fetching page %d' % page_num,
+                                          data=json.dumps(queryObject).encode('utf-8'),
+                                          headers={'Content-Type': 'text/plain'})
+            if results['err'] is not None:
+                raise ExtractorError('API returned an error: %s' % results['err'][0])
+            entries.extend(self._extract_playlist_entries(results['result']['results']))
+
+            meta = results['result']['queryInfo']
+            if len(entries) >= n:
+                entries = entries[:n]
+                break
+            elif meta['resultCount'] == 0:
+                break
+
+        common_topic = None
+        if entries:
+            common_topic = entries[0]['series']
+            for entry in entries:
+                common_topic = common_topic if entry['series'] == common_topic else None
+
+        return self.playlist_result(entries, playlist_title=common_topic or query)
+
+
+class MediathekViewWebIE(InfoExtractor):
+    # @see https://github.com/mediathekview/mediathekviewweb
+    IE_NAME = 'mediathekviewweb'
+    _VALID_URL = r'https?://mediathekviewweb\.de/\#query=(?P<id>.+)'
+
+    _TESTS = [
+        {
+            # Test for everywhere.
+            'url': 'https://mediathekviewweb.de/#query=!ard%20%23Tagesschau%2020%2CUhr&everywhere=true',
+            'info_dict': {
+                'title': '!ard #Tagesschau 20,Uhr',
+            },
+            # Without everywhere, there are <100 results.
+            'playlist_mincount': 365,
+            'params': {
+                'skip_download': True,
+            },
+        },
+        {
+            # Test for non-future videos.
+            'url': 'https://mediathekviewweb.de/#query=%23sport%2Cim%2Costen%20biathlon&future=false',
+            'info_dict': {
+                'title': 'Sport im Osten',
+            },
+            # Future yields 4 results instead.
+            'playlist_maxcount': 2,
+            'params': {
+                'skip_download': True,
+            },
+        },
+    ]
+
+    def _real_extract(self, url):
+        query_hash = self._match_id(url)
+
+        url_stub = '?query=' + query_hash
+        query = compat_parse_qs(compat_urlparse.urlparse(url_stub).query)
+        search = query['query'][0]
+        query.pop('query')
+
+        if len(query) > 0:
+            # Detect global flags, MVW is very strict about accepted values.
+            extractor = MediathekViewWebSearchIE(self._downloader)
+            if query.get('everywhere', []) == ['true']:
+                extractor._everywhere = True
+            if query.get('future', []) == ['false']:
+                extractor._future = False
+            return extractor._real_extract('mvwsearchall:' + search)
+
+        return self.url_result('mvwsearchall:' + search, ie=MediathekViewWebSearchIE.ie_key())

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')