Merge d2d6b49a5a into e1b3fa242c

Fixes #32877

test/test_youtube_signature.py

@@ -166,6 +166,14 @@ _NSIG_TESTS = [
         'https://www.youtube.com/s/player/b22ef6e7/player_ias.vflset/en_US/base.js',
         'b6HcntHGkvBLk_FRf', 'kNPW6A7FyP2l8A',
     ),
+    (
+        'https://www.youtube.com/s/player/3400486c/player_ias.vflset/en_US/base.js',
+        'lL46g3XifCKUZn1Xfw', 'z767lhet6V2Skl',
+    ),
+    (
+        'https://www.youtube.com/s/player/5604538d/player_ias.vflset/en_US/base.js',
+        '7X-he4jjvMx7BCX', 'sViSydX8IHtdWA',
+    ),
 ]
 
 

youtube_dl/extractor/redtube.py

@@ -1,33 +1,49 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
 import re
 
 from .common import InfoExtractor
+from ..compat import compat_str
 from ..utils import (
     determine_ext,
     ExtractorError,
     int_or_none,
+    join_nonempty,
     merge_dicts,
     str_to_int,
+    T,
+    traverse_obj,
     unified_strdate,
     url_or_none,
+    urljoin,
 )
 
 
 class RedTubeIE(InfoExtractor):
     _VALID_URL = r'https?://(?:(?:\w+\.)?redtube\.com/|embed\.redtube\.com/\?.*?\bid=)(?P<id>[0-9]+)'
+    _EMBED_REGEX = [r'<iframe[^>]+?src=["\'](?P<url>(?:https?:)?//embed\.redtube\.com/\?.*?\bid=\d+)']
     _TESTS = [{
-        'url': 'http://www.redtube.com/66418',
-        'md5': 'fc08071233725f26b8f014dba9590005',
+        'url': 'https://www.redtube.com/38864951',
+        'md5': 'd7de9cb32e8adb3f6379f1a30f655fae',
         'info_dict': {
-            'id': '66418',
+            'id': '38864951',
             'ext': 'mp4',
-            'title': 'Sucked on a toilet',
-            'upload_date': '20110811',
-            'duration': 596,
+            'title': 'Public Sex on the Balcony in Freezing Paris! Amateur Couple LeoLulu',
+            'description': 'Watch video Public Sex on the Balcony in Freezing Paris! Amateur Couple LeoLulu on Redtube, home of free Blowjob porn videos and Blonde sex movies online. Video length: (10:46) - Uploaded by leolulu - Verified User - Starring Pornstar: LeoLulu',
+            'upload_date': '20210111',
+            'timestamp': 1610343109,
+            'duration': 646,
             'view_count': int,
             'age_limit': 18,
-        }
+            'thumbnail': r're:https://\wi-ph\.rdtcdn\.com/videos/.+/.+\.jpg',
+        },
+        'expected_warnings': [
+            'Failed to download m3u8 information: HTTP Error 404',
+        ],
+        'params': {
+            'format': '[format_id !^= hls]',
+        },
     }, {
         'url': 'http://embed.redtube.com/?bgcolor=000000&id=1443286',
         'only_matching': True,
@@ -36,11 +52,11 @@ class RedTubeIE(InfoExtractor):
         'only_matching': True,
     }]
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return re.findall(
-            r'<iframe[^>]+?src=["\'](?P<url>(?:https?:)?//embed\.redtube\.com/\?.*?\bid=\d+)',
-            webpage)
+    @classmethod
+    def _extract_urls(cls, webpage):
+        for embed_re in cls._EMBED_REGEX:
+            for from_ in re.findall(embed_re, webpage):
+                yield from_
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
@@ -71,40 +87,53 @@ class RedTubeIE(InfoExtractor):
             self._search_regex(
                 r'sources\s*:\s*({.+?})', webpage, 'source', default='{}'),
             video_id, fatal=False)
-        if sources and isinstance(sources, dict):
-            for format_id, format_url in sources.items():
-                if format_url:
-                    formats.append({
-                        'url': format_url,
-                        'format_id': format_id,
-                        'height': int_or_none(format_id),
-                    })
-        medias = self._parse_json(
-            self._search_regex(
-                r'mediaDefinition["\']?\s*:\s*(\[.+?}\s*\])', webpage,
-                'media definitions', default='{}'),
-            video_id, fatal=False)
-        if medias and isinstance(medias, list):
-            for media in medias:
-                format_url = url_or_none(media.get('videoUrl'))
+
+        def full_url(u):
+            return urljoin(url, u)
+
+        for fmt in traverse_obj(sources, (T(dict.items), {
+                'url': (1, T(full_url)),
+                'format_id': (2, T(compat_str)),
+                'height': (2, T(int_or_none)), })):
+            if 'url' in fmt:
+                formats.append(fmt)
+
+        medias = self._search_regex(
+            r'''mediaDefinitions?["']?\s*:\s*(\[[\s\S]+?}\s*\])''', webpage,
+            'media definitions', default='{}')
+        medias = self._parse_json(medias, video_id, fatal=False)
+        for fmt in traverse_obj(medias, (Ellipsis, T(dict))):
+            format_url = full_url(fmt.get('videoUrl'))
+            if not format_url:
+                continue
+            more_media = None
+            if fmt['format'] == 'hls' or (fmt['format'] == 'mp4' and not fmt.get('quality')):
+                more_media = self._download_json(format_url, video_id, fatal=False)
+            if more_media is None:
+                more_media = [fmt]
+            for fmt in traverse_obj(more_media, (Ellipsis, {
+                    'url': ('videoUrl', T(full_url)),
+                    'ext': ('format', T(compat_str)),
+                    'format_id': ('quality', T(compat_str)), })):
+                format_url = fmt.get('url')
                 if not format_url:
                     continue
-                if media.get('format') == 'hls' or determine_ext(format_url) == 'm3u8':
+                if fmt.get('ext') == 'hls' or determine_ext(format_url) == 'm3u8':
                     formats.extend(self._extract_m3u8_formats(
                         format_url, video_id, 'mp4',
                         entry_protocol='m3u8_native', m3u8_id='hls',
                         fatal=False))
                     continue
-                format_id = media.get('quality')
-                formats.append({
-                    'url': format_url,
-                    'format_id': format_id,
-                    'height': int_or_none(format_id),
-                })
+                fmt['height'] = int_or_none(fmt.get('format_id'))
+                fmt['format_id'] = join_nonempty('ext', 'format_id', from_dict=fmt)
+                formats.append(fmt)
         if not formats:
-            video_url = self._html_search_regex(
-                r'<source src="(.+?)" type="video/mp4">', webpage, 'video URL')
-            formats.append({'url': video_url})
+            video_url = url_or_none(self._html_search_regex(
+                r'<source src="(.+?)" type="video/mp4">', webpage, 'video URL'))
+            if video_url:
+                formats.append({'url': video_url})
+
+        self._check_formats(formats, video_id)
         self._sort_formats(formats)
 
         thumbnail = self._og_search_thumbnail(webpage)

youtube_dl/extractor/youtube.py

@@ -1647,7 +1647,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         except JSInterpreter.Exception as e:
             self.report_warning(
                 '%s (%s %s)' % (
-                    'Unable to decode n-parameter: download likely to be throttled',
+                    'Unable to decode n-parameter: expect download to be blocked or throttled',
                     error_to_compat_str(e),
                     traceback.format_exc()),
                 video_id=video_id)
@@ -1659,18 +1659,22 @@ 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)
             r'''(?x)
-                (?:\(\s*(?P<b>[a-z])\s*=\s*String\s*\.\s*fromCharCode\s*\(\s*110\s*\)\s*,(?P<c>[a-z])\s*=\s*[a-z]\s*)?
-                \.\s*get\s*\(\s*(?(b)(?P=b)|"n")(?:\s*\)){2}\s*&&\s*\(\s*(?(c)(?P=c)|b)\s*=\s*
+                (?:\(\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'))
         if not idx:
             return func_name
 
         return self._parse_json(self._search_regex(
-            r'var {0}\s*=\s*(\[.+?\])\s*[,;]'.format(re.escape(func_name)), jscode,
+            r'var\s+{0}\s*=\s*(\[.+?\])\s*[,;]'.format(re.escape(func_name)), jscode,
             'Initial JS player n function list ({0}.{1})'.format(func_name, idx)),
             func_name, transform_source=js_to_json)[int(idx)]