]> jfr.im git - yt-dlp.git/blobdiff - yt_dlp/extractor/generic.py
[extractor/generic] Attempt to detect live HLS (#6775)
[yt-dlp.git] / yt_dlp / extractor / generic.py
index 5da77273d804886f5ad3f73979112b895f6628ef..87cf11d6bdef84c163a63b390bca86d810b066df 100644 (file)
@@ -14,7 +14,9 @@
     ExtractorError,
     UnsupportedError,
     determine_ext,
+    determine_protocol,
     dict_get,
+    extract_basic_auth,
     format_field,
     int_or_none,
     is_html,
@@ -31,7 +33,9 @@
     unescapeHTML,
     unified_timestamp,
     unsmuggle_url,
+    update_url_query,
     url_or_none,
+    urljoin,
     variadic,
     xpath_attr,
     xpath_text,
@@ -864,21 +868,7 @@ class GenericIE(InfoExtractor):
             },
         },
         {
-            # JWPlayer config passed as variable
-            'url': 'http://www.txxx.com/videos/3326530/ariele/',
-            'info_dict': {
-                'id': '3326530_hq',
-                'ext': 'mp4',
-                'title': 'ARIELE | Tube Cup',
-                'uploader': 'www.txxx.com',
-                'age_limit': 18,
-            },
-            'params': {
-                'skip_download': True,
-            }
-        },
-        {
-            # Video.js embed, multiple formats
+            # Youtube embed, formerly: Video.js embed, multiple formats
             'url': 'http://ortcam.com/solidworks-урок-6-настройка-чертежа_33f9b7351.html',
             'info_dict': {
                 'id': 'yygqldloqIk',
@@ -905,6 +895,7 @@ class GenericIE(InfoExtractor):
             'params': {
                 'skip_download': True,
             },
+            'skip': '404 Not Found',
         },
         # rtl.nl embed
         {
@@ -1547,19 +1538,6 @@ class GenericIE(InfoExtractor):
             },
             'add_ie': ['WashingtonPost'],
         },
-        {
-            # Mediaset embed
-            'url': 'http://www.tgcom24.mediaset.it/politica/serracchiani-voglio-vivere-in-una-societa-aperta-reazioni-sproporzionate-_3071354-201702a.shtml',
-            'info_dict': {
-                'id': '720642',
-                'ext': 'mp4',
-                'title': 'Serracchiani: "Voglio vivere in una società aperta, con tutela del patto di fiducia"',
-            },
-            'params': {
-                'skip_download': True,
-            },
-            'add_ie': ['Mediaset'],
-        },
         {
             # JOJ.sk embeds
             'url': 'https://www.noviny.sk/slovensko/238543-slovenskom-sa-prehnala-vlna-silnych-burok',
@@ -1864,11 +1842,6 @@ class GenericIE(InfoExtractor):
                 'title': 'I AM BIO Podcast | BIO',
             },
             'playlist_mincount': 52,
-        },
-        {
-            # Sibnet embed (https://help.sibnet.ru/?sibnet_video_embed)
-            'url': 'https://phpbb3.x-tk.ru/bbcode-video-sibnet-t24.html',
-            'only_matching': True,
         }, {
             # WimTv embed player
             'url': 'http://www.msmotor.tv/wearefmi-pt-2-2021/',
@@ -1885,11 +1858,13 @@ class GenericIE(InfoExtractor):
                 'display_id': 'kelis-4th-of-july',
                 'ext': 'mp4',
                 'title': 'Kelis - 4th Of July',
-                'thumbnail': 'https://kvs-demo.com/contents/videos_screenshots/0/105/preview.jpg',
+                'description': 'Kelis - 4th Of July',
+                'thumbnail': r're:https://(?:www\.)?kvs-demo.com/contents/videos_screenshots/0/105/preview.jpg',
             },
             'params': {
                 'skip_download': True,
             },
+            'expected_warnings': ['Untested major version'],
         }, {
             # KVS Player
             'url': 'https://www.kvs-demo.com/embed/105/',
@@ -1898,35 +1873,12 @@ class GenericIE(InfoExtractor):
                 'display_id': 'kelis-4th-of-july',
                 'ext': 'mp4',
                 'title': 'Kelis - 4th Of July / Embed Player',
-                'thumbnail': 'https://kvs-demo.com/contents/videos_screenshots/0/105/preview.jpg',
+                'thumbnail': r're:https://(?:www\.)?kvs-demo.com/contents/videos_screenshots/0/105/preview.jpg',
             },
             'params': {
                 'skip_download': True,
             },
         }, {
-            # KVS Player
-            'url': 'https://thisvid.com/videos/french-boy-pantsed/',
-            'md5': '3397979512c682f6b85b3b04989df224',
-            'info_dict': {
-                'id': '2400174',
-                'display_id': 'french-boy-pantsed',
-                'ext': 'mp4',
-                'title': 'French Boy Pantsed - ThisVid.com',
-                'thumbnail': 'https://media.thisvid.com/contents/videos_screenshots/2400000/2400174/preview.mp4.jpg',
-            }
-        }, {
-            # KVS Player
-            'url': 'https://thisvid.com/embed/2400174/',
-            'md5': '3397979512c682f6b85b3b04989df224',
-            'info_dict': {
-                'id': '2400174',
-                'display_id': 'french-boy-pantsed',
-                'ext': 'mp4',
-                'title': 'French Boy Pantsed - ThisVid.com',
-                'thumbnail': 'https://media.thisvid.com/contents/videos_screenshots/2400000/2400174/preview.mp4.jpg',
-            }
-        }, {
-            # KVS Player
             'url': 'https://youix.com/video/leningrad-zoj/',
             'md5': '94f96ba95706dc3880812b27b7d8a2b8',
             'info_dict': {
@@ -1934,8 +1886,8 @@ class GenericIE(InfoExtractor):
                 'display_id': 'leningrad-zoj',
                 'ext': 'mp4',
                 'title': 'Клип: Ленинград - ЗОЖ скачать, смотреть онлайн | Youix.com',
-                'thumbnail': 'https://youix.com/contents/videos_screenshots/18000/18485/preview_480x320_youix_com.mp4.jpg',
-            }
+                'thumbnail': r're:https://youix.com/contents/videos_screenshots/18000/18485/preview(?:_480x320_youix_com.mp4)?\.jpg',
+            },
         }, {
             # KVS Player
             'url': 'https://youix.com/embed/18485',
@@ -1945,19 +1897,20 @@ class GenericIE(InfoExtractor):
                 'display_id': 'leningrad-zoj',
                 'ext': 'mp4',
                 'title': 'Ленинград - ЗОЖ',
-                'thumbnail': 'https://youix.com/contents/videos_screenshots/18000/18485/preview_480x320_youix_com.mp4.jpg',
-            }
+                'thumbnail': r're:https://youix.com/contents/videos_screenshots/18000/18485/preview(?:_480x320_youix_com.mp4)?\.jpg',
+            },
         }, {
             # KVS Player
             'url': 'https://bogmedia.org/videos/21217/40-nochey-40-nights-2016/',
             'md5': '94166bdb26b4cb1fb9214319a629fc51',
             'info_dict': {
                 'id': '21217',
-                'display_id': '40-nochey-40-nights-2016',
+                'display_id': '40-nochey-2016',
                 'ext': 'mp4',
                 'title': '40 ночей (2016) - BogMedia.org',
+                'description': 'md5:4e6d7d622636eb7948275432eb256dc3',
                 'thumbnail': 'https://bogmedia.org/contents/videos_screenshots/21000/21217/preview_480p.mp4.jpg',
-            }
+            },
         },
         {
             # KVS Player (for sites that serve kt_player.js via non-https urls)
@@ -1967,9 +1920,9 @@ class GenericIE(InfoExtractor):
                 'id': '389508',
                 'display_id': 'syren-de-mer-onlyfans-05-07-2020have-a-happy-safe-holiday5f014e68a220979bdb8cd-source',
                 'ext': 'mp4',
-                'title': 'Syren De Mer  onlyfans_05-07-2020Have_a_happy_safe_holiday5f014e68a220979bdb8cd_source / Embed плеер',
-                'thumbnail': 'http://www.camhub.world/contents/videos_screenshots/389000/389508/preview.mp4.jpg',
-            }
+                'title': 'Syren De Mer onlyfans_05-07-2020Have_a_happy_safe_holiday5f014e68a220979bdb8cd_source / Embed плеер',
+                'thumbnail': r're:https?://www\.camhub\.world/contents/videos_screenshots/389000/389508/preview\.mp4\.jpg',
+            },
         },
         {
             # Reddit-hosted video that will redirect and be processed by RedditIE
@@ -2172,7 +2125,79 @@ class GenericIE(InfoExtractor):
                 'age_limit': 0,
                 'direct': True,
             }
-        }
+        },
+        {
+            'note': 'server returns data in brotli compression by default if `accept-encoding: *` is specified.',
+            'url': 'https://www.extra.cz/cauky-lidi-70-dil-babis-predstavil-pohadky-prymulanek-nebo-andrejovy-nove-saty-ac867',
+            'info_dict': {
+                'id': 'cauky-lidi-70-dil-babis-predstavil-pohadky-prymulanek-nebo-andrejovy-nove-saty-ac867',
+                'ext': 'mp4',
+                'title': 'čauky lidi 70 finall',
+                'description': 'čauky lidi 70 finall',
+                'thumbnail': 'h',
+                'upload_date': '20220606',
+                'timestamp': 1654513791,
+                'duration': 318.0,
+                'direct': True,
+                'age_limit': 0,
+            },
+        },
+        {
+            'note': 'JW Player embed with unicode-escape sequences in URL',
+            'url': 'https://www.medici.tv/en/concerts/lahav-shani-mozart-mahler-israel-philharmonic-abu-dhabi-classics',
+            'info_dict': {
+                'id': 'm',
+                'ext': 'mp4',
+                'title': 'Lahav Shani conducts the Israel Philharmonic\'s first-ever concert in Abu Dhabi',
+                'description': 'Mahler\'s ',
+                'uploader': 'www.medici.tv',
+                'age_limit': 0,
+                'thumbnail': r're:^https?://.+\.jpg',
+            },
+            'params': {
+                'skip_download': True,
+            },
+        },
+        {
+            'url': 'https://shooshtime.com/videos/284002/just-out-of-the-shower-joi/',
+            'md5': 'e2f0a4c329f7986280b7328e24036d60',
+            'info_dict': {
+                'id': '284002',
+                'display_id': 'just-out-of-the-shower-joi',
+                'ext': 'mp4',
+                'title': 'Just Out Of The Shower JOI - Shooshtime',
+                'thumbnail': 'https://i.shoosh.co/contents/videos_screenshots/284000/284002/preview.mp4.jpg',
+                'height': 720,
+                'age_limit': 18,
+            },
+        },
+        {
+            'note': 'Live HLS direct link',
+            'url': 'https://d18j67ugtrocuq.cloudfront.net/out/v1/2767aec339144787926bd0322f72c6e9/index.m3u8',
+            'info_dict': {
+                'id': 'index',
+                'title': r're:index',
+                'ext': 'mp4',
+                'live_status': 'is_live',
+            },
+            'params': {
+                'skip_download': 'm3u8',
+            },
+        },
+        {
+            'note': 'Video.js VOD HLS',
+            'url': 'https://gist.githubusercontent.com/bashonly/2aae0862c50f4a4b84f220c315767208/raw/e3380d413749dabbe804c9c2d8fd9a45142475c7/videojs_hls_test.html',
+            'info_dict': {
+                'id': 'videojs_hls_test',
+                'title': 'video',
+                'ext': 'mp4',
+                'age_limit': 0,
+                'duration': 1800,
+            },
+            'params': {
+                'skip_download': 'm3u8',
+            },
+        },
     ]
 
     def report_following_redirect(self, new_url):
@@ -2189,6 +2214,42 @@ def report_detected(self, name, num=1, note=None):
 
         self._downloader.write_debug(f'Identified {num} {name}{format_field(note, None, "; %s")}')
 
+    def _extra_manifest_info(self, info, manifest_url):
+        fragment_query = self._configuration_arg('fragment_query', [None], casesense=True)[0]
+        if fragment_query is not None:
+            info['extra_param_to_segment_url'] = (
+                urllib.parse.urlparse(fragment_query).query or fragment_query
+                or urllib.parse.urlparse(manifest_url).query or None)
+
+        hex_or_none = lambda x: x if re.fullmatch(r'(0x)?[\da-f]+', x, re.IGNORECASE) else None
+        info['hls_aes'] = traverse_obj(self._configuration_arg('hls_key', casesense=True), {
+            'uri': (0, {url_or_none}), 'key': (0, {hex_or_none}), 'iv': (1, {hex_or_none}),
+        }) or None
+
+        variant_query = self._configuration_arg('variant_query', [None], casesense=True)[0]
+        if variant_query is not None:
+            query = urllib.parse.parse_qs(
+                urllib.parse.urlparse(variant_query).query or variant_query
+                or urllib.parse.urlparse(manifest_url).query)
+            for fmt in self._downloader._get_formats(info):
+                fmt['url'] = update_url_query(fmt['url'], query)
+
+        # Attempt to detect live HLS or set VOD duration
+        m3u8_format = next((f for f in self._downloader._get_formats(info)
+                            if determine_protocol(f) == 'm3u8_native'), None)
+        if m3u8_format:
+            is_live = self._configuration_arg('is_live', [None])[0]
+            if is_live is not None:
+                info['live_status'] = 'not_live' if is_live == 'false' else 'is_live'
+                return
+            headers = m3u8_format.get('http_headers') or info.get('http_headers')
+            duration = self._extract_m3u8_vod_duration(
+                m3u8_format['url'], info.get('id'), note='Checking m3u8 live status',
+                errnote='Failed to download m3u8 media playlist', headers=headers)
+            if not duration:
+                info['live_status'] = 'is_live'
+            info['duration'] = info.get('duration') or duration
+
     def _extract_rss(self, url, video_id, doc):
         NS_MAP = {
             'itunes': 'http://www.itunes.com/dtds/podcast-1.0.dtd',
@@ -2231,43 +2292,87 @@ def itunes(key):
             'entries': entries,
         }
 
-    def _kvs_getrealurl(self, video_url, license_code):
+    @classmethod
+    def _kvs_get_real_url(cls, video_url, license_code):
         if not video_url.startswith('function/0/'):
             return video_url  # not obfuscated
 
-        url_path, _, url_query = video_url.partition('?')
-        urlparts = url_path.split('/')[2:]
-        license = self._kvs_getlicensetoken(license_code)
-        newmagic = urlparts[5][:32]
+        parsed = urllib.parse.urlparse(video_url[len('function/0/'):])
+        license = cls._kvs_get_license_token(license_code)
+        urlparts = parsed.path.split('/')
 
-        for o in range(len(newmagic) - 1, -1, -1):
-            new = ''
-            l = (o + sum(int(n) for n in license[o:])) % 32
+        HASH_LENGTH = 32
+        hash = urlparts[3][:HASH_LENGTH]
+        indices = list(range(HASH_LENGTH))
 
-            for i in range(0, len(newmagic)):
-                if i == o:
-                    new += newmagic[l]
-                elif i == l:
-                    new += newmagic[o]
-                else:
-                    new += newmagic[i]
-            newmagic = new
+        # Swap indices of hash according to the destination calculated from the license token
+        accum = 0
+        for src in reversed(range(HASH_LENGTH)):
+            accum += license[src]
+            dest = (src + accum) % HASH_LENGTH
+            indices[src], indices[dest] = indices[dest], indices[src]
+
+        urlparts[3] = ''.join(hash[index] for index in indices) + urlparts[3][HASH_LENGTH:]
+        return urllib.parse.urlunparse(parsed._replace(path='/'.join(urlparts)))
 
-        urlparts[5] = newmagic + urlparts[5][32:]
-        return '/'.join(urlparts) + '?' + url_query
+    @staticmethod
+    def _kvs_get_license_token(license):
+        license = license.replace('$', '')
+        license_values = [int(char) for char in license]
 
-    def _kvs_getlicensetoken(self, license):
-        modlicense = license.replace('$', '').replace('0', '1')
-        center = int(len(modlicense) / 2)
+        modlicense = license.replace('0', '1')
+        center = len(modlicense) // 2
         fronthalf = int(modlicense[:center + 1])
         backhalf = int(modlicense[center:])
+        modlicense = str(4 * abs(fronthalf - backhalf))[:center + 1]
+
+        return [
+            (license_values[index + offset] + current) % 10
+            for index, current in enumerate(map(int, modlicense))
+            for offset in range(4)
+        ]
+
+    def _extract_kvs(self, url, webpage, video_id):
+        flashvars = self._search_json(
+            r'(?s:<script\b[^>]*>.*?var\s+flashvars\s*=)',
+            webpage, 'flashvars', video_id, transform_source=js_to_json)
+
+        # extract the part after the last / as the display_id from the
+        # canonical URL.
+        display_id = self._search_regex(
+            r'(?:<link href="https?://[^"]+/(.+?)/?" rel="canonical"\s*/?>'
+            r'|<link rel="canonical" href="https?://[^"]+/(.+?)/?"\s*/?>)',
+            webpage, 'display_id', fatal=False)
+        title = self._html_search_regex(r'<(?:h1|title)>(?:Video: )?(.+?)</(?:h1|title)>', webpage, 'title')
 
-        modlicense = str(4 * abs(fronthalf - backhalf))
-        retval = ''
-        for o in range(0, center + 1):
-            for i in range(1, 5):
-                retval += str((int(license[o + i]) + int(modlicense[o])) % 10)
-        return retval
+        thumbnail = flashvars['preview_url']
+        if thumbnail.startswith('//'):
+            protocol, _, _ = url.partition('/')
+            thumbnail = protocol + thumbnail
+
+        url_keys = list(filter(re.compile(r'^video_(?:url|alt_url\d*)$').match, flashvars.keys()))
+        formats = []
+        for key in url_keys:
+            if '/get_file/' not in flashvars[key]:
+                continue
+            format_id = flashvars.get(f'{key}_text', key)
+            formats.append({
+                'url': urljoin(url, self._kvs_get_real_url(flashvars[key], flashvars['license_code'])),
+                'format_id': format_id,
+                'ext': 'mp4',
+                **(parse_resolution(format_id) or parse_resolution(flashvars[key])),
+                'http_headers': {'Referer': url},
+            })
+            if not formats[-1].get('height'):
+                formats[-1]['quality'] = 1
+
+        return {
+            'id': flashvars['video_id'],
+            'display_id': display_id,
+            'title': title,
+            'thumbnail': thumbnail,
+            'formats': formats,
+        }
 
     def _real_extract(self, url):
         if url.startswith('//'):
@@ -2323,13 +2428,12 @@ def _real_extract(self, url):
         # It may probably better to solve this by checking Content-Type for application/octet-stream
         # after a HEAD request, but not sure if we can rely on this.
         full_response = self._request_webpage(url, video_id, headers={
-            'Accept-Encoding': '*',
+            'Accept-Encoding': 'identity',
             **smuggled_data.get('http_headers', {})
         })
         new_url = full_response.geturl()
-        if new_url == urllib.parse.urlparse(url)._replace(scheme='https').geturl():
-            url = new_url
-        elif url != new_url:
+        url = urllib.parse.urlparse(url)._replace(scheme=urllib.parse.urlparse(new_url).scheme).geturl()
+        if new_url != extract_basic_auth(url)[0]:
             self.report_following_redirect(new_url)
             if force_videoid:
                 new_url = smuggle_url(new_url, {'force_videoid': force_videoid})
@@ -2348,12 +2452,13 @@ def _real_extract(self, url):
             self.report_detected('direct video link')
             headers = smuggled_data.get('http_headers', {})
             format_id = str(m.group('format_id'))
+            ext = determine_ext(url)
             subtitles = {}
-            if format_id.endswith('mpegurl'):
+            if format_id.endswith('mpegurl') or ext == 'm3u8':
                 formats, subtitles = self._extract_m3u8_formats_and_subtitles(url, video_id, 'mp4', headers=headers)
-            elif format_id.endswith('mpd') or format_id.endswith('dash+xml'):
+            elif format_id.endswith('mpd') or format_id.endswith('dash+xml') or ext == 'mpd':
                 formats, subtitles = self._extract_mpd_formats_and_subtitles(url, video_id, headers=headers)
-            elif format_id == 'f4m':
+            elif format_id == 'f4m' or ext == 'f4m':
                 formats = self._extract_f4m_formats(url, video_id, headers=headers)
             else:
                 formats = [{
@@ -2365,8 +2470,9 @@ def _real_extract(self, url):
             info_dict.update({
                 'formats': formats,
                 'subtitles': subtitles,
-                'http_headers': headers,
+                'http_headers': headers or None,
             })
+            self._extra_manifest_info(info_dict, url)
             return info_dict
 
         if not self.get_param('test', False) and not is_intentional:
@@ -2379,6 +2485,7 @@ def _real_extract(self, url):
         if first_bytes.startswith(b'#EXTM3U'):
             self.report_detected('M3U playlist')
             info_dict['formats'], info_dict['subtitles'] = self._extract_m3u8_formats_and_subtitles(url, video_id, 'mp4')
+            self._extra_manifest_info(info_dict, url)
             return info_dict
 
         # Maybe it's a direct link to a video?
@@ -2429,6 +2536,7 @@ def _real_extract(self, url):
                     doc,
                     mpd_base_url=full_response.geturl().rpartition('/')[0],
                     mpd_url=url)
+                self._extra_manifest_info(info_dict, url)
                 self.report_detected('DASH manifest')
                 return info_dict
             elif re.match(r'^{http://ns\.adobe\.com/f4m/[12]\.0}manifest$', doc.tag):
@@ -2517,8 +2625,7 @@ def _extract_embeds(self, url, webpage, *, urlh=None, info_dict={}):
             varname = mobj.group(1)
             sources = variadic(self._parse_json(
                 mobj.group(2), video_id, transform_source=js_to_json, fatal=False) or [])
-            formats = []
-            subtitles = {}
+            formats, subtitles, src = [], {}, None
             for source in sources:
                 src = source.get('src')
                 if not src or not isinstance(src, str):
@@ -2541,7 +2648,8 @@ def _extract_embeds(self, url, webpage, *, urlh=None, info_dict={}):
                         m3u8_id='hls', fatal=False)
                     formats.extend(fmts)
                     self._merge_subtitles(subs, target=subtitles)
-                else:
+
+                if not formats:
                     formats.append({
                         'url': src,
                         'ext': (mimetype2ext(src_type)
@@ -2555,11 +2663,11 @@ def _extract_embeds(self, url, webpage, *, urlh=None, info_dict={}):
             for sub_match in re.finditer(rf'(?s){re.escape(varname)}' r'\.addRemoteTextTrack\(({.+?})\s*,\s*(?:true|false)\)', webpage):
                 sub = self._parse_json(
                     sub_match.group(1), video_id, transform_source=js_to_json, fatal=False) or {}
-                src = str_or_none(sub.get('src'))
-                if not src:
+                sub_src = str_or_none(sub.get('src'))
+                if not sub_src:
                     continue
                 subtitles.setdefault(dict_get(sub, ('language', 'srclang')) or 'und', []).append({
-                    'url': urllib.parse.urljoin(url, src),
+                    'url': urllib.parse.urljoin(url, sub_src),
                     'name': sub.get('label'),
                     'http_headers': {
                         'Referer': actual_url,
@@ -2567,7 +2675,21 @@ def _extract_embeds(self, url, webpage, *, urlh=None, info_dict={}):
                 })
             if formats or subtitles:
                 self.report_detected('video.js embed')
-                return [{'formats': formats, 'subtitles': subtitles}]
+                info_dict = {'formats': formats, 'subtitles': subtitles}
+                if formats:
+                    self._extra_manifest_info(info_dict, src)
+                return [info_dict]
+
+        # Look for generic KVS player (before json-ld bc of some urls that break otherwise)
+        found = self._search_regex((
+            r'<script\b[^>]+?\bsrc\s*=\s*(["\'])https?://(?:(?!\1)[^?#])+/kt_player\.js\?v=(?P<ver>\d+(?:\.\d+)+)\1[^>]*>',
+            r'kt_player\s*\(\s*(["\'])(?:(?!\1)[\w\W])+\1\s*,\s*(["\'])https?://(?:(?!\2)[^?#])+/kt_player\.swf\?v=(?P<ver>\d+(?:\.\d+)+)\2\s*,',
+        ), webpage, 'KVS player', group='ver', default=False)
+        if found:
+            self.report_detected('KVS Player')
+            if found.split('.')[0] not in ('4', '5', '6'):
+                self.report_warning(f'Untested major version ({found}) in player engine - download may fail.')
+            return [self._extract_kvs(url, webpage, video_id)]
 
         # Looking for http://schema.org/VideoObject
         json_ld = self._search_json_ld(webpage, video_id, default={})
@@ -2611,52 +2733,6 @@ def filter_video(urls):
                 ['"]?file['"]?\s*:\s*["\'](.*?)["\']''', webpage))
             if found:
                 self.report_detected('JW Player embed')
-        if not found:
-            # Look for generic KVS player
-            found = re.search(r'<script [^>]*?src="https?://.+?/kt_player\.js\?v=(?P<ver>(?P<maj_ver>\d+)(\.\d+)+)".*?>', webpage)
-            if found:
-                self.report_detected('KWS Player')
-                if found.group('maj_ver') not in ['4', '5']:
-                    self.report_warning('Untested major version (%s) in player engine--Download may fail.' % found.group('ver'))
-                flashvars = re.search(r'(?ms)<script.*?>.*?var\s+flashvars\s*=\s*(\{.*?\});.*?</script>', webpage)
-                flashvars = self._parse_json(flashvars.group(1), video_id, transform_source=js_to_json)
-
-                # extract the part after the last / as the display_id from the
-                # canonical URL.
-                display_id = self._search_regex(
-                    r'(?:<link href="https?://[^"]+/(.+?)/?" rel="canonical"\s*/?>'
-                    r'|<link rel="canonical" href="https?://[^"]+/(.+?)/?"\s*/?>)',
-                    webpage, 'display_id', fatal=False
-                )
-                title = self._html_search_regex(r'<(?:h1|title)>(?:Video: )?(.+?)</(?:h1|title)>', webpage, 'title')
-
-                thumbnail = flashvars['preview_url']
-                if thumbnail.startswith('//'):
-                    protocol, _, _ = url.partition('/')
-                    thumbnail = protocol + thumbnail
-
-                url_keys = list(filter(re.compile(r'video_url|video_alt_url\d*').fullmatch, flashvars.keys()))
-                formats = []
-                for key in url_keys:
-                    if '/get_file/' not in flashvars[key]:
-                        continue
-                    format_id = flashvars.get(f'{key}_text', key)
-                    formats.append({
-                        'url': self._kvs_getrealurl(flashvars[key], flashvars['license_code']),
-                        'format_id': format_id,
-                        'ext': 'mp4',
-                        **(parse_resolution(format_id) or parse_resolution(flashvars[key]))
-                    })
-                    if not formats[-1].get('height'):
-                        formats[-1]['quality'] = 1
-
-                return [{
-                    'id': flashvars['video_id'],
-                    'display_id': display_id,
-                    'title': title,
-                    'thumbnail': thumbnail,
-                    'formats': formats,
-                }]
         if not found:
             # Broaden the search a little bit
             found = filter_video(re.findall(r'[^A-Za-z0-9]?(?:file|source)=(http[^\'"&]*)', webpage))
@@ -2737,6 +2813,7 @@ def filter_video(urls):
 
         entries = []
         for video_url in orderedSet(found):
+            video_url = video_url.encode().decode('unicode-escape')
             video_url = unescapeHTML(video_url)
             video_url = video_url.replace('\\/', '/')
             video_url = urllib.parse.urljoin(url, video_url)
@@ -2776,8 +2853,10 @@ def filter_video(urls):
                 return [self._extract_xspf_playlist(video_url, video_id)]
             elif ext == 'm3u8':
                 entry_info_dict['formats'], entry_info_dict['subtitles'] = self._extract_m3u8_formats_and_subtitles(video_url, video_id, ext='mp4', headers=headers)
+                self._extra_manifest_info(entry_info_dict, video_url)
             elif ext == 'mpd':
                 entry_info_dict['formats'], entry_info_dict['subtitles'] = self._extract_mpd_formats_and_subtitles(video_url, video_id, headers=headers)
+                self._extra_manifest_info(entry_info_dict, video_url)
             elif ext == 'f4m':
                 entry_info_dict['formats'] = self._extract_f4m_formats(video_url, video_id, headers=headers)
             elif re.search(r'(?i)\.(?:ism|smil)/manifest', video_url) and video_url != url: