Merge b6bfdb25f2 into c5098961b0

Now also succeeds with player b12cc44b

docs/supportedsites.md

@@ -910,6 +910,9 @@
  - **SVTPage**
  - **SVTPlay**: SVT Play and Öppet arkiv
  - **SVTSeries**
+ - **switchtube**
+ - **switchtube:channel**
+ - **switchtube:profile**
  - **SWRMediathek**
  - **Syfy**
  - **SztvHu**

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

@@ -1235,6 +1235,11 @@ from .svt import (
     SVTPlayIE,
     SVTSeriesIE,
 )
+from .switchtube import (
+  SwitchTubeIE,
+  SwitchTubeProfileIE,
+  SwitchTubeChannelIE,
+)
 from .swrmediathek import SWRMediathekIE
 from .syfy import SyfyIE
 from .sztvhu import SztvHuIE

youtube_dl/extractor/switchtube.py

@@ -0,0 +1,219 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import itertools
+import re
+
+from .common import InfoExtractor
+from ..utils import (
+    clean_html,
+    extract_attributes,
+    float_or_none,
+    get_element_by_attribute,
+    int_or_none,
+    parse_iso8601,
+    strip_or_none,
+    urljoin,
+)
+
+
+class SwitchTubeIE(InfoExtractor):
+    _VALID_URL = r'https?://tube\.switch\.ch/videos/(?P<id>[\da-zA-Z]+)'
+    IE_NAME = 'switchtube'
+    _TESTS = [{
+        'url': 'https://tube.switch.ch/videos/0T1XfaIFSX',
+        'info_dict': {
+            'id': '0T1XfaIFSX',
+            'title': '2016_ASE_sqC03-Entretien',
+            'channel': 'ASE Assistant-e socio-éducatif-ve CFC',
+            'channel_url': 'https://tube.switch.ch/channels/bsaer76yoL',
+            'channel_id': 'bsaer76yoL',
+            'ext': 'mp4',
+            'description': None,
+            'thumbnail': r're:^https?://tube.switch.ch/image/representations/[\w-]+$',
+            'license': 'All rights reserved',
+            'creator': 'Jean-Marc Pouly from Eidgenössische Hochschule für Berufsbildung',
+            'uploader': 'Jean-Marc Pouly from Eidgenössische Hochschule für Berufsbildung',
+            'uploader_url': 'https://tube.switch.ch/profiles/42481',
+            'uploader_id': '42481',
+            'upload_date': '20220309',
+            'timestamp': 1646839068,
+        }
+    }, {
+        'url': 'https://tube.switch.ch/videos/0cf3886d',
+        'info_dict': {
+            'id': '0cf3886d',
+            'ext': 'mp4',
+            'title': 'Introduction: Mini-Batches in On- and Off-Policy Deep Reinforcement Learning',
+            'license': 'All rights reserved',
+            'description': 'One of the challenges in Deep Reinforcement Learning is to decorrelate the data. How this is possible with replay buffers is explained here.',
+            'thumbnail': r're:^https?://tube.switch.ch/image/representations/[\w-]+$',
+            'channel': 'CS-456 Artificial Neural Networks',
+            'channel_url': 'https://tube.switch.ch/channels/1deb03e0',
+            'channel_id': '1deb03e0',
+            'timestamp': 1590733406,
+            'upload_date': '20200529',
+            'creator': 'Wulfram Gerstner from École polytechnique fédérale de Lausanne (EPFL)',
+            'uploader': 'Wulfram Gerstner from École polytechnique fédérale de Lausanne (EPFL)',
+            'uploader_url': 'https://tube.switch.ch/profiles/94248',
+            'uploader_id': '94248',
+        }
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+        title = self._html_search_regex(r'(?s)<title\b[^>]*>(.+?)</title>', webpage, 'title')
+
+        info = {
+            'id': video_id,
+            'title': title,
+            'is_live': False,
+        }
+
+        info['view_count'] = int_or_none(self._html_search_regex(r'(?s)<p\b[^>]*>\s*(\d+)\s+views?\s*</p>', webpage,
+                                                                 'view_count', default=None))
+        info['license'] = self._html_search_regex(r'''(?s)<span\b[^>]*?\bproperty\s*=\s*["'](?:\b[cd]c:license\b\s*){2,}[^>]+>(.+)</span>''',
+                                                  webpage, 'license', default=None)
+        info['description'] = strip_or_none(clean_html(
+            get_element_by_attribute('property', 'dc:description', webpage)))
+        info['duration'] = float_or_none(
+            self._search_regex(r'''\bdata-duration\s*=\s*["']([\d.]+)''',
+                               webpage, 'duration', default=None))
+
+        channel_groups = self._search_regex(r'''(?s)<a\b[^>]+?\bhref\s*=\s*["']/channels/(.{4,}?)</a>''',
+                                            webpage, 'channel groups', default='')
+        channel_groups = re.split(r'''(?s)["'][^>]*>\s*''', channel_groups, 1)
+        if len(channel_groups) == 2:
+            for i, ch in enumerate(('channel_id', 'channel')):
+                info[ch] = strip_or_none(channel_groups[i])
+            if info['channel_id']:
+                info['channel_url'] = 'https://tube.switch.ch/channels/' + info['channel_id']
+
+        def outer_elements_by_attribute(attr, value, html, tag=None, escape_value=True, include_inner=False):
+            """generate matching HTML element strings
+               if include_inner, tuples of (element, content)"""
+            pattern = r'''(?s)(?P<element><(%s)\b[^>]+?\b%s\s*=\s*("|')%s\b[^>]+>)(?P<inner>.*?)</\2>''' % \
+                (re.escape(tag) if tag is not None else r'\w+', attr, re.escape(value) if escape_value else value)
+            matches = re.finditer(pattern, html)
+            for m in matches:
+                yield m.group('element', 'inner') if include_inner else m.group('element')
+
+        for dt in outer_elements_by_attribute('property', 'dc:date', webpage, tag='span'):
+            dt = extract_attributes(dt)
+            if dt.get('class') == 'dt-published':
+                info['timestamp'] = parse_iso8601(dt.get('content'))
+                break
+
+        creator_groups = self._search_regex(
+            r'''(?s)<span\b[^>]+?\bclass\s*=\s*("|')(?:(?!\1).)*\bp-author\b(?:(?!\1).)*\1\s*property\s*=\s*["']dc:creator\b[^>]+>\s*(.*?<span\b[^>]+?\bclass\s*=\s*["']p-name\b.*</span>).*?</span>''',
+            webpage, 'creator groups', default='', group=2)
+        creator_groups = re.match(r'''(?s)<a\b[^>]+?\bhref\s*=\s*("|')/profiles/(?P<profile_id>.+?)\1[^>]*>\s*<span\b[^>]+>\s*(?P<creator_name>.+?)\s*</span>[,\s]*<span\b[^>]+\bclass\s*=\s*["']p-organization-name\b[^>]+>\s*(?P<organization_name>.+?)\s*</span>''', creator_groups)
+        if creator_groups:
+            creator_groups = creator_groups.groupdict()
+            info['uploader'] = info['creator'] = ' from '.join((creator_groups['creator_name'], creator_groups['organization_name']))
+            info['uploader_id'] = creator_groups['profile_id']
+            info['uploader_url'] = 'https://tube.switch.ch/profiles/' + info['uploader_id']
+
+        parsed_media_entries = self._parse_html5_media_entries(url, webpage, video_id)[0]
+        info['thumbnail'] = parsed_media_entries['thumbnail']
+        info['formats'] = parsed_media_entries['formats']
+        self._sort_formats(info['formats'])
+
+        return info
+
+
+class SwitchTubeProfileIE(InfoExtractor):
+    _VALID_URL = r'https?://tube\.switch\.ch/profiles/(?P<id>[\da-zA-Z]+)'
+    IE_NAME = 'switchtube:profile'
+    _TESTS = [{
+        'url': 'https://tube.switch.ch/profiles/94248',
+        'info_dict': {
+            'id': '94248',
+            'title': 'Wulfram Gerstner',
+            'description': None,
+        },
+        'playlist_mincount': 94,
+    }]
+
+    @classmethod
+    def suitable(cls, url):
+        return False if SwitchTubeIE.suitable(url) else super(SwitchTubeProfileIE, cls).suitable(url)
+
+    def _real_extract(self, url):
+        channel_id = self._match_id(url)
+        webpage = self._download_webpage(url, channel_id)
+        channel_title = self._html_search_regex(r'(?s)<title\b[^>]*>(.+?)</title>', webpage, 'title', default=None)
+
+        description = self._html_search_regex(r'''(?s)<div\b[^>]+class\s*=\s*("|')p-summary formatted\1[^>]+property\s*=\s*("|')dc:description\2[^>]*>\s*(.+?)\s*</div>''',
+                                              webpage, 'description', default=None, group=3)
+
+        entries = []
+        next_page = None
+        for current_page_number in itertools.count(1):
+            if next_page:
+                webpage = self._download_webpage(next_page, channel_id, note='Downloading page %d' % (current_page_number, ))
+
+            for _, video_path, video_id, title in re.findall(
+                    r'''(?s)<a\b[^>]+\bhref\s*=\s*("|')(/videos/((?:(?!\1).)+?))\1[^>]*>\s*<div\b[^>]+\bclass\s*=\s*["']title\b[^>]+>(.+?)</div>''',
+                    webpage):
+                video_url = urljoin(url, video_path)
+                if video_url:
+                    entries.append(self.url_result(video_url, ie=SwitchTubeIE.ie_key(), video_id=video_id))
+
+            next_page = self._search_regex(
+                r'''<a\b[^>]+?\bhref\s*=\s*("|')(?P<path>/profiles/%s\?(?:(?!\1).)+)\1[^>]*>\s*Next\s*</a>''' % (channel_id,),
+                webpage, 'next page', group='path', default=None)
+            if next_page:
+                next_page = urljoin(url, next_page)
+            if not next_page:
+                break
+
+        return self.playlist_result(entries, channel_id, channel_title,
+                                    description)
+
+
+class SwitchTubeChannelIE(InfoExtractor):
+    _VALID_URL = r'https?://tube\.switch\.ch/channels/(?P<id>[\da-zA-Z]+)'
+    IE_NAME = 'switchtube:channel'
+    _TESTS = [{
+        'url': 'https://tube.switch.ch/channels/1deb03e0',
+        'info_dict': {
+            'id': '1deb03e0',
+            'title': 'CS-456 Artificial Neural Networks',
+            'description': 'Class on Artificial Neural Networks and Reinforcement Learning designed for EPFL master students in CS and related disciplines.'
+        },
+        'playlist_mincount': 94,
+    }]
+
+    @classmethod
+    def suitable(cls, url):
+        return False if SwitchTubeIE.suitable(url) else super(SwitchTubeChannelIE, cls).suitable(url)
+
+    def _real_extract(self, url):
+        channel_id = self._match_id(url)
+
+        entries = []
+        for current_page_number in itertools.count(0):
+            page_url = urljoin(url, '/channels/%s?order=episodes&page=%d' % (channel_id, current_page_number))
+            webpage = self._download_webpage(page_url, channel_id, note='Downloading page %d' % (current_page_number + 1, ))
+
+            if current_page_number == 0:
+                channel_title = self._html_search_regex(r'(?s)<title\b[^>]*>(.+?)</title>', webpage, 'title', default=None)
+                description = self._html_search_regex(r'''(?s)<div\b[^>]+class\s*=\s*("|')description formatted\1[^>]*>\s*<p>\s*(.+?)\s*</p>\s*</div>''',
+                                                      webpage, 'description', default=None, group=2)
+
+            this_page_still_has_something = False
+            for _, video_path, video_id in re.findall(
+                    r'''(?s)<a\b[^>]+\bhref\s*=\s*("|')(/videos/((?:(?!\1).)+?))\1[^>]*>''',
+                    webpage):
+                video_url = urljoin(url, video_path)
+                if video_url:
+                    this_page_still_has_something = True
+                    entries.append(self.url_result(video_url, ie=SwitchTubeIE.ie_key(), video_id=video_id))
+
+            if not this_page_still_has_something:
+                break
+
+        return self.playlist_result(entries, channel_id, channel_title,
+                                    description)

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
-            # old: .get("n"))&&(b=nfunc[idx](b)
-            # older: .get("n"))&&(b=nfunc(b)
+            # or:  (b="nn"[+a.D],c=a.get(b))&&(c=nfunc[idx](c)
+            # or:  (PL(a),b=a.j.n||null)&&(b=nfunc[idx](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*(?:
-                    String\s*\.\s*fromCharCode\s*\(\s*110\s*\)|
-                    "n+"\[\s*\+?s*[\w$.]+\s*]
-                )\s*,(?P<c>[a-z])\s*=\s*[a-z]\s*)?
-                \.\s*get\s*\(\s*(?(b)(?P=b)|"n{1,2}")(?:\s*\)){2}\s*&&\s*\(\s*(?(c)(?P=c)|b)\s*=\s*
-                (?P<nfunc>[a-zA-Z_$][\w$]*)(?:\s*\[(?P<idx>\d+)\])?\s*\(\s*[\w$]+\s*\)
-            ''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'))
+                \((?:[\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$.]+\s*]
+                    )\s*(?:,[\w$()\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'),
+            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(len(argvals) == 1, 'takes exactly one argument')
-                    return obj[argvals[0]:]
+                    assertion(isinstance(obj, (list, compat_str)), 'must be applied on a list or string')
+                    # From [1]:
+                    # .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')