Merge b3ab97c05e into e0727e4ab6

Fixes 21792b88b7 (commitcomment-140705274), thx: vonProteus

devscripts/zsh-completion.in

@@ -3,26 +3,8 @@
 __youtube_dl() {
     local curcontext="$curcontext" fileopts diropts cur prev
     typeset -A opt_args
-    fileopts="{{fileopts}}"
-    diropts="{{diropts}}"
-    cur=$words[CURRENT]
-    case $cur in
-        :)
-            _arguments '*: :(::ytfavorites ::ytrecommended ::ytsubscriptions ::ytwatchlater ::ythistory)'
-        ;;
-        *)
-            prev=$words[CURRENT-1]
-            if [[ ${prev} =~ ${fileopts} ]]; then
-                _path_files
-            elif [[ ${prev} =~ ${diropts} ]]; then
-                _path_files -/
-            elif [[ ${prev} == "--recode-video" ]]; then
-                _arguments '*: :(mp4 flv ogg webm mkv)'
-            else
-                _arguments '*: :({{flags}})'
-            fi
-        ;;
-    esac
+    _arguments {{args}} \
+      '*: :(::ytfavorites ::ytrecommended ::ytsubscriptions ::ytwatchlater ::ythistory)'
 }
 
-__youtube_dl
+__youtube_dl

devscripts/zsh-completion.py

@@ -4,6 +4,7 @@ from __future__ import unicode_literals
 import os
 from os.path import dirname as dirn
 import sys
+import optparse
 
 sys.path.insert(0, dirn(dirn((os.path.abspath(__file__)))))
 import youtube_dl
@@ -17,30 +18,44 @@ ZSH_COMPLETION_TEMPLATE = "devscripts/zsh-completion.in"
 def build_completion(opt_parser):
     opts = [opt for group in opt_parser.option_groups
             for opt in group.option_list]
-    opts_file = [opt for opt in opts if opt.metavar == "FILE"]
-    opts_dir = [opt for opt in opts if opt.metavar == "DIR"]
 
-    fileopts = []
-    for opt in opts_file:
+    # escaping is hard:
+    #  - help may contain colons
+    #  - metavar must have colons : escaped
+    #  - single quotes must be removed
+    #  - help must have square brackets [] escaped
+    def metaparse(opt):
+        if "--recode-video" == opt.get_opt_string():
+            return ":{}:(mp4 flv ogg webm mkv)".format(opt.metavar)
+        if opt.metavar is None:
+            return ""
+        if opt.metavar == "FILE":
+            return ":FILE:_files"
+        if opt.metavar == "DIR":
+            return ":DIR:_directories"
+        else:
+            return ":{}:".format(opt.metavar.replace(":", "\\:"))
+
+    def helpescape(opthelp):
+        if opthelp == optparse.SUPPRESS_HELP:
+            return ""
+        return "[{}]".format(opthelp.replace("'", "\"").replace("]", "\\]").replace("[", "\\["))
+
+    def optionexclude(opt):
+        # When an argument has a long and short version, the arguments entry shall be
+        # _arguments \
+        #   "(-t --thing)"{-t,--thing}"[do things]:WHAT_THING:"
+        # i.e. in parentheses with space the explanation of redundancy and in curly braces
+        # regular shell expansion to create two mostly identical entries.
         if opt._short_opts:
-            fileopts.extend(opt._short_opts)
-        if opt._long_opts:
-            fileopts.extend(opt._long_opts)
+            return "({0} {1})'{{{0},{1}}}'".format(opt._short_opts[0], opt.get_opt_string())
+        return "{}".format(opt.get_opt_string())
 
-    diropts = []
-    for opt in opts_dir:
-        if opt._short_opts:
-            diropts.extend(opt._short_opts)
-        if opt._long_opts:
-            diropts.extend(opt._long_opts)
-
-    flags = [opt.get_opt_string() for opt in opts]
+    mytest = ["'{}{}{}'".format(optionexclude(opt), helpescape(opt.help), metaparse(opt)) for opt in opts]
 
     template = read_file(ZSH_COMPLETION_TEMPLATE)
 
-    template = template.replace("{{fileopts}}", "|".join(fileopts))
-    template = template.replace("{{diropts}}", "|".join(diropts))
-    template = template.replace("{{flags}}", " ".join(flags))
+    template = template.replace("{{args}}", " \\\n      ".join(mytest))
 
     write_file(ZSH_COMPLETION_FILE, template)
 

test/test_InfoExtractor.py

@@ -993,7 +993,8 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
                     'tbr': 5997.485,
                     'width': 1920,
                     'height': 1080,
-                }]
+                }],
+                {},
             ), (
                 # https://github.com/ytdl-org/youtube-dl/pull/14844
                 'urls_only',
@@ -1076,7 +1077,8 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
                     'tbr': 4400,
                     'width': 1920,
                     'height': 1080,
-                }]
+                }],
+                {},
             ), (
                 # https://github.com/ytdl-org/youtube-dl/issues/20346
                 # Media considered unfragmented even though it contains
@@ -1122,18 +1124,185 @@ jwplayer("mediaplayer").setup({"abouttext":"Visit Indie DB","aboutlink":"http:\/
                     'width': 360,
                     'height': 360,
                     'fps': 30,
-                }]
+                }],
+                {},
+            ), (
+                # https://github.com/ytdl-org/youtube-dl/issues/30235
+                # Bento4 generated test mpd
+                # mp4dash --mpd-name=manifest.mpd --no-split --use-segment-list mediafiles
+                'url_and_range',
+                'http://unknown/manifest.mpd',  # mpd_url
+                'http://unknown/',  # mpd_base_url
+                [{
+                    'manifest_url': 'http://unknown/manifest.mpd',
+                    'fragment_base_url': 'http://unknown/',
+                    'ext': 'm4a',
+                    'format_id': 'audio-und-mp4a.40.2',
+                    'format_note': 'DASH audio',
+                    'container': 'm4a_dash',
+                    'protocol': 'http_dash_segments',
+                    'acodec': 'mp4a.40.2',
+                    'vcodec': 'none',
+                    'tbr': 98.808,
+                }, {
+                    'manifest_url': 'http://unknown/manifest.mpd',
+                    'fragment_base_url': 'http://unknown/',
+                    'ext': 'mp4',
+                    'format_id': 'video-avc1',
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'protocol': 'http_dash_segments',
+                    'acodec': 'none',
+                    'vcodec': 'avc1.4D401E',
+                    'tbr': 699.597,
+                    'width': 768,
+                    'height': 432
+                }],
+                {},
+            ), (
+                # https://github.com/ytdl-org/youtube-dl/issues/27575
+                # GPAC generated test mpd
+                # MP4Box -dash 10000 -single-file -out manifest.mpd mediafiles
+                'range_only',
+                'http://unknown/manifest.mpd',  # mpd_url
+                'http://unknown/',  # mpd_base_url
+                [{
+                    'manifest_url': 'http://unknown/manifest.mpd',
+                    'fragment_base_url': 'http://unknown/audio_dashinit.mp4',
+                    'ext': 'm4a',
+                    'format_id': '2',
+                    'format_note': 'DASH audio',
+                    'container': 'm4a_dash',
+                    'protocol': 'http_dash_segments',
+                    'acodec': 'mp4a.40.2',
+                    'vcodec': 'none',
+                    'tbr': 98.096,
+                }, {
+                    'manifest_url': 'http://unknown/manifest.mpd',
+                    'fragment_base_url': 'http://unknown/video_dashinit.mp4',
+                    'ext': 'mp4',
+                    'format_id': '1',
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'protocol': 'http_dash_segments',
+                    'acodec': 'none',
+                    'vcodec': 'avc1.4D401E',
+                    'tbr': 526.987,
+                    'width': 768,
+                    'height': 432
+                }],
+                {},
+            ), (
+                'subtitles',
+                'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/',
+                [{
+                    'format_id': 'audio=128001',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'm4a',
+                    'tbr': 128.001,
+                    'asr': 48000,
+                    'format_note': 'DASH audio',
+                    'container': 'm4a_dash',
+                    'vcodec': 'none',
+                    'acodec': 'mp4a.40.2',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }, {
+                    'format_id': 'video=100000',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'mp4',
+                    'width': 336,
+                    'height': 144,
+                    'tbr': 100,
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'vcodec': 'avc1.4D401F',
+                    'acodec': 'none',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }, {
+                    'format_id': 'video=326000',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'mp4',
+                    'width': 562,
+                    'height': 240,
+                    'tbr': 326,
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'vcodec': 'avc1.4D401F',
+                    'acodec': 'none',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }, {
+                    'format_id': 'video=698000',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'mp4',
+                    'width': 844,
+                    'height': 360,
+                    'tbr': 698,
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'vcodec': 'avc1.4D401F',
+                    'acodec': 'none',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }, {
+                    'format_id': 'video=1493000',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'mp4',
+                    'width': 1126,
+                    'height': 480,
+                    'tbr': 1493,
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'vcodec': 'avc1.4D401F',
+                    'acodec': 'none',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }, {
+                    'format_id': 'video=4482000',
+                    'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'ext': 'mp4',
+                    'width': 1688,
+                    'height': 720,
+                    'tbr': 4482,
+                    'format_note': 'DASH video',
+                    'container': 'mp4_dash',
+                    'vcodec': 'avc1.4D401F',
+                    'acodec': 'none',
+                    'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                    'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                    'protocol': 'http_dash_segments',
+                }],
+                {
+                    'en': [
+                        {
+                            'ext': 'mp4',
+                            'manifest_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                            'url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/manifest.mpd',
+                            'fragment_base_url': 'https://sdn-global-streaming-cache-3qsdn.akamaized.net/stream/3144/files/17/07/672975/3144-kZT4LWMQw6Rh7Kpd.ism/dash/',
+                            'protocol': 'http_dash_segments',
+                        }
+                    ]
+                },
             )
         ]
 
-        for mpd_file, mpd_url, mpd_base_url, expected_formats in _TEST_CASES:
+        for mpd_file, mpd_url, mpd_base_url, expected_formats, expected_subtitles in _TEST_CASES:
             with open('./test/testdata/mpd/%s.mpd' % mpd_file,
                       mode='r', encoding='utf-8') as f:
-                formats = self.ie._parse_mpd_formats(
+                formats, subtitles = self.ie._parse_mpd_formats_and_subtitles(
                     compat_etree_fromstring(f.read().encode('utf-8')),
                     mpd_base_url=mpd_base_url, mpd_url=mpd_url)
                 self.ie._sort_formats(formats)
                 expect_value(self, formats, expected_formats, None)
+                expect_value(self, subtitles, expected_subtitles, None)
 
     def test_parse_f4m_formats(self):
         _TEST_CASES = [

test/test_compat.py

@@ -23,6 +23,7 @@ from youtube_dl.compat import (
     compat_urllib_parse_unquote,
     compat_urllib_parse_unquote_plus,
     compat_urllib_parse_urlencode,
+    compat_urllib_request,
 )
 
 
@@ -135,6 +136,19 @@ class TestCompat(unittest.TestCase):
         self.assertEqual(compat_casefold('\u03a3'), '\u03c3')
         self.assertEqual(compat_casefold('A\u0345\u03a3'), 'a\u03b9\u03c3')
 
+    def test_compat_urllib_request_Request(self):
+        self.assertEqual(
+            compat_urllib_request.Request('http://127.0.0.1', method='PUT').get_method(),
+            'PUT')
+
+        class PUTrequest(compat_urllib_request.Request):
+            def get_method(self):
+                return 'PUT'
+
+        self.assertEqual(
+            PUTrequest('http://127.0.0.1').get_method(),
+            'PUT')
+
 
 if __name__ == '__main__':
     unittest.main()

test/test_downloader_external.py

@@ -18,6 +18,7 @@ from test.helper import (
 )
 from youtube_dl import YoutubeDL
 from youtube_dl.compat import (
+    compat_contextlib_suppress,
     compat_http_cookiejar_Cookie,
     compat_http_server,
     compat_kwargs,
@@ -35,6 +36,9 @@ from youtube_dl.downloader.external import (
     HttpieFD,
     WgetFD,
 )
+from youtube_dl.postprocessor import (
+    FFmpegPostProcessor,
+)
 import threading
 
 TEST_SIZE = 10 * 1024
@@ -227,7 +231,17 @@ class TestAria2cFD(unittest.TestCase):
             self.assertIn('--load-cookies=%s' % downloader._cookies_tempfile, cmd)
 
 
-@ifExternalFDAvailable(FFmpegFD)
+# Handle delegated availability
+def ifFFmpegFDAvailable(externalFD):
+    # raise SkipTest, or set False!
+    avail = ifExternalFDAvailable(externalFD) and False
+    with compat_contextlib_suppress(Exception):
+        avail = FFmpegPostProcessor(downloader=None).available
+    return unittest.skipUnless(
+        avail, externalFD.get_basename() + ' not found')
+
+
+@ifFFmpegFDAvailable(FFmpegFD)
 class TestFFmpegFD(unittest.TestCase):
     _args = []
 

test/test_subtitles.py

@@ -295,6 +295,7 @@ class TestNRKSubtitles(BaseTestSubtitles):
     def test_allsubtitles(self):
         self.DL.params['writesubtitles'] = True
         self.DL.params['allsubtitles'] = True
+        self.DL.params['format'] = 'best/bestvideo'
         subtitles = self.getSubtitles()
         self.assertEqual(set(subtitles.keys()), set(['nb-ttv']))
         self.assertEqual(md5(subtitles['nb-ttv']), '67e06ff02d0deaf975e68f6cb8f6a149')

test/test_utils.py

@@ -81,6 +81,7 @@ from youtube_dl.utils import (
     sanitize_filename,
     sanitize_path,
     sanitize_url,
+    sanitized_Request,
     shell_quote,
     smuggle_url,
     str_or_none,
@@ -255,6 +256,18 @@ class TestUtil(unittest.TestCase):
         self.assertEqual(sanitize_url('https://foo.bar'), 'https://foo.bar')
         self.assertEqual(sanitize_url('foo bar'), 'foo bar')
 
+    def test_sanitized_Request(self):
+        self.assertFalse(sanitized_Request('http://foo.bar').has_header('Authorization'))
+        self.assertFalse(sanitized_Request('http://:foo.bar').has_header('Authorization'))
+        self.assertEqual(sanitized_Request('http://@foo.bar').get_header('Authorization'),
+                         'Basic Og==')
+        self.assertEqual(sanitized_Request('http://:pass@foo.bar').get_header('Authorization'),
+                         'Basic OnBhc3M=')
+        self.assertEqual(sanitized_Request('http://user:@foo.bar').get_header('Authorization'),
+                         'Basic dXNlcjo=')
+        self.assertEqual(sanitized_Request('http://user:pass@foo.bar').get_header('Authorization'),
+                         'Basic dXNlcjpwYXNz')
+
     def test_expand_path(self):
         def env(var):
             return '%{0}%'.format(var) if sys.platform == 'win32' else '${0}'.format(var)
@@ -512,11 +525,14 @@ class TestUtil(unittest.TestCase):
         self.assertEqual(float_or_none(set()), None)
 
     def test_int_or_none(self):
+        self.assertEqual(int_or_none(42), 42)
         self.assertEqual(int_or_none('42'), 42)
         self.assertEqual(int_or_none(''), None)
         self.assertEqual(int_or_none(None), None)
         self.assertEqual(int_or_none([]), None)
         self.assertEqual(int_or_none(set()), None)
+        self.assertEqual(int_or_none('42', base=8), 34)
+        self.assertRaises(TypeError, int_or_none(42, base=8))
 
     def test_str_to_int(self):
         self.assertEqual(str_to_int('123,456'), 123456)

test/testdata/mpd/range_only.mpd

@@ -0,0 +1,35 @@
+<?xml version="1.0"?>
+<!-- MPD file Generated with GPAC version 1.0.1-revrelease at 2021-11-27T20:53:11.690Z -->
+<MPD xmlns="urn:mpeg:dash:schema:mpd:2011" minBufferTime="PT1.500S" type="static" mediaPresentationDuration="PT0H0M30.196S" maxSegmentDuration="PT0H0M10.027S" profiles="urn:mpeg:dash:profile:full:2011">
+ <ProgramInformation moreInformationURL="http://gpac.io">
+  <Title>manifest.mpd generated by GPAC</Title>
+ </ProgramInformation>
+
+ <Period duration="PT0H0M30.196S">
+  <AdaptationSet segmentAlignment="true" maxWidth="768" maxHeight="432" maxFrameRate="30000/1001" par="16:9" lang="und" startWithSAP="1">
+   <Representation id="1" mimeType="video/mp4" codecs="avc1.4D401E" width="768" height="432" frameRate="30000/1001" sar="1:1" bandwidth="526987">
+    <BaseURL>video_dashinit.mp4</BaseURL>
+    <SegmentList timescale="90000" duration="900000">
+     <Initialization range="0-881"/>
+     <SegmentURL mediaRange="882-876094" indexRange="882-925"/>
+     <SegmentURL mediaRange="876095-1466732" indexRange="876095-876138"/>
+     <SegmentURL mediaRange="1466733-1953615" indexRange="1466733-1466776"/>
+     <SegmentURL mediaRange="1953616-1994211" indexRange="1953616-1953659"/>
+    </SegmentList>
+   </Representation>
+  </AdaptationSet>
+  <AdaptationSet segmentAlignment="true" lang="und" startWithSAP="1">
+   <Representation id="2" mimeType="audio/mp4" codecs="mp4a.40.2" audioSamplingRate="48000" bandwidth="98096">
+    <AudioChannelConfiguration schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011" value="2"/>
+    <BaseURL>audio_dashinit.mp4</BaseURL>
+    <SegmentList timescale="48000" duration="480000">
+     <Initialization range="0-752"/>
+     <SegmentURL mediaRange="753-124129" indexRange="753-796"/>
+     <SegmentURL mediaRange="124130-250544" indexRange="124130-124173"/>
+     <SegmentURL mediaRange="250545-374929" indexRange="250545-250588"/>
+    </SegmentList>
+   </Representation>
+  </AdaptationSet>
+ </Period>
+</MPD>
+

test/testdata/mpd/subtitles.mpd

@@ -0,0 +1,351 @@
+<?xml version="1.0" encoding="utf-8"?>
+<!-- Created with Unified Streaming Platform (version=1.10.18-20255) -->
+<MPD
+  xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
+  xmlns="urn:mpeg:dash:schema:mpd:2011"
+  xsi:schemaLocation="urn:mpeg:dash:schema:mpd:2011 http://standards.iso.org/ittf/PubliclyAvailableStandards/MPEG-DASH_schema_files/DASH-MPD.xsd"
+  type="static"
+  mediaPresentationDuration="PT14M48S"
+  maxSegmentDuration="PT1M"
+  minBufferTime="PT10S"
+  profiles="urn:mpeg:dash:profile:isoff-live:2011">
+  <Period
+    id="1"
+    duration="PT14M48S">
+    <BaseURL>dash/</BaseURL>
+    <AdaptationSet
+      id="1"
+      group="1"
+      contentType="audio"
+      segmentAlignment="true"
+      audioSamplingRate="48000"
+      mimeType="audio/mp4"
+      codecs="mp4a.40.2"
+      startWithSAP="1">
+      <AudioChannelConfiguration
+        schemeIdUri="urn:mpeg:dash:23003:3:audio_channel_configuration:2011"
+        value="2" />
+      <Role schemeIdUri="urn:mpeg:dash:role:2011" value="main" />
+      <SegmentTemplate
+        timescale="48000"
+        initialization="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$.dash"
+        media="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$-$Time$.dash">
+        <SegmentTimeline>
+          <S t="0" d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="96256" r="2" />
+          <S d="95232" />
+          <S d="3584" />
+        </SegmentTimeline>
+      </SegmentTemplate>
+      <Representation
+        id="audio=128001"
+        bandwidth="128001">
+      </Representation>
+    </AdaptationSet>
+    <AdaptationSet
+      id="2"
+      group="3"
+      contentType="text"
+      lang="en"
+      mimeType="application/mp4"
+      codecs="stpp"
+      startWithSAP="1">
+      <Role schemeIdUri="urn:mpeg:dash:role:2011" value="subtitle" />
+      <SegmentTemplate
+        timescale="1000"
+        initialization="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$.dash"
+        media="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$-$Time$.dash">
+        <SegmentTimeline>
+          <S t="0" d="60000" r="9" />
+          <S d="24000" />
+        </SegmentTimeline>
+      </SegmentTemplate>
+      <Representation
+        id="textstream_eng=1000"
+        bandwidth="1000">
+      </Representation>
+    </AdaptationSet>
+    <AdaptationSet
+      id="3"
+      group="2"
+      contentType="video"
+      par="960:409"
+      minBandwidth="100000"
+      maxBandwidth="4482000"
+      maxWidth="1689"
+      maxHeight="720"
+      segmentAlignment="true"
+      mimeType="video/mp4"
+      codecs="avc1.4D401F"
+      startWithSAP="1">
+      <Role schemeIdUri="urn:mpeg:dash:role:2011" value="main" />
+      <SegmentTemplate
+        timescale="12288"
+        initialization="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$.dash"
+        media="3144-kZT4LWMQw6Rh7Kpd-$RepresentationID$-$Time$.dash">
+        <SegmentTimeline>
+          <S t="0" d="24576" r="443" />
+        </SegmentTimeline>
+      </SegmentTemplate>
+      <Representation
+        id="video=100000"
+        bandwidth="100000"
+        width="336"
+        height="144"
+        sar="2880:2863"
+        scanType="progressive">
+      </Representation>
+      <Representation
+        id="video=326000"
+        bandwidth="326000"
+        width="562"
+        height="240"
+        sar="115200:114929"
+        scanType="progressive">
+      </Representation>
+      <Representation
+        id="video=698000"
+        bandwidth="698000"
+        width="844"
+        height="360"
+        sar="86400:86299"
+        scanType="progressive">
+      </Representation>
+      <Representation
+        id="video=1493000"
+        bandwidth="1493000"
+        width="1126"
+        height="480"
+        sar="230400:230267"
+        scanType="progressive">
+      </Representation>
+      <Representation
+        id="video=4482000"
+        bandwidth="4482000"
+        width="1688"
+        height="720"
+        sar="86400:86299"
+        scanType="progressive">
+      </Representation>
+    </AdaptationSet>
+  </Period>
+</MPD>

test/testdata/mpd/url_and_range.mpd

@@ -0,0 +1,32 @@
+<?xml version="1.0" ?>
+<MPD xmlns="urn:mpeg:dash:schema:mpd:2011" profiles="urn:mpeg:dash:profile:isoff-live:2011" minBufferTime="PT10.01S" mediaPresentationDuration="PT30.097S" type="static">
+  <!-- Created with Bento4 mp4-dash.py, VERSION=2.0.0-639 -->
+  <Period>
+    <!-- Video -->
+    <AdaptationSet mimeType="video/mp4" segmentAlignment="true" startWithSAP="1" maxWidth="768" maxHeight="432">
+      <Representation id="video-avc1" codecs="avc1.4D401E" width="768" height="432" scanType="progressive" frameRate="30000/1001" bandwidth="699597">
+        <SegmentList timescale="1000" duration="10010">
+          <Initialization sourceURL="video-frag.mp4" range="36-746"/>
+          <SegmentURL media="video-frag.mp4" mediaRange="747-876117"/>
+          <SegmentURL media="video-frag.mp4" mediaRange="876118-1466913"/>
+          <SegmentURL media="video-frag.mp4" mediaRange="1466914-1953954"/>
+          <SegmentURL media="video-frag.mp4" mediaRange="1953955-1994652"/>
+        </SegmentList>
+      </Representation>
+    </AdaptationSet>
+    <!-- Audio -->
+    <AdaptationSet mimeType="audio/mp4" startWithSAP="1" segmentAlignment="true">
+      <Representation id="audio-und-mp4a.40.2" codecs="mp4a.40.2" bandwidth="98808" audioSamplingRate="48000">
+        <AudioChannelConfiguration schemeIdUri="urn:mpeg:mpegB:cicp:ChannelConfiguration" value="2"/>
+        <SegmentList timescale="1000" duration="10010">
+          <Initialization sourceURL="audio-frag.mp4" range="32-623"/>
+          <SegmentURL media="audio-frag.mp4" mediaRange="624-124199"/>
+          <SegmentURL media="audio-frag.mp4" mediaRange="124200-250303"/>
+          <SegmentURL media="audio-frag.mp4" mediaRange="250304-374365"/>
+          <SegmentURL media="audio-frag.mp4" mediaRange="374366-374836"/>
+        </SegmentList>
+      </Representation>
+    </AdaptationSet>
+  </Period>
+</MPD>
+

youtube_dl/YoutubeDL.py

@@ -2635,12 +2635,12 @@ class YoutubeDL(object):
             self.to_screen(msg('[info] %s is already present', label.title()))
             return 'exists'
         else:
-            self.to_screen(msg('[info] Writing %s as JSON to: ' + infofn, label))
+            self.to_screen(msg('[info] Writing %s as JSON to: ', label) + infofn)
             try:
                 write_json_file(self.filter_requested_info(info_dict), infofn)
                 return True
             except (OSError, IOError):
-                self.report_error(msg('Cannot write %s to JSON file ' + infofn, label))
+                self.report_error(msg('Cannot write %s to JSON file ', label) + infofn)
                 return
 
     def _write_thumbnails(self, info_dict, filename):

youtube_dl/compat.py

@@ -58,19 +58,26 @@ except ImportError:  # Python 2
 
 # Also fix up lack of method arg in old Pythons
 try:
-    _req = compat_urllib_request.Request
-    _req('http://127.0.0.1', method='GET')
+    type(compat_urllib_request.Request('http://127.0.0.1', method='GET'))
 except TypeError:
-    class _request(object):
-        def __new__(cls, url, *args, **kwargs):
-            method = kwargs.pop('method', None)
-            r = _req(url, *args, **kwargs)
-            if method:
-                r.get_method = types.MethodType(lambda _: method, r)
-            return r
+    def _add_init_method_arg(cls):
 
-    compat_urllib_request.Request = _request
+        init = cls.__init__
 
+        def wrapped_init(self, *args, **kwargs):
+            method = kwargs.pop('method', 'GET')
+            init(self, *args, **kwargs)
+            if any(callable(x.__dict__.get('get_method')) for x in (self.__class__, self) if x != cls):
+                # allow instance or its subclass to override get_method()
+                return
+            if self.has_data() and method == 'GET':
+                method = 'POST'
+            self.get_method = types.MethodType(lambda _: method, self)
+
+        cls.__init__ = wrapped_init
+
+    _add_init_method_arg(compat_urllib_request.Request)
+    del _add_init_method_arg
 
 try:
     import urllib.error as compat_urllib_error
@@ -2414,29 +2421,26 @@ except ImportError:  # Python 2
 compat_urllib_request_urlretrieve = compat_urlretrieve
 
 try:
+    from HTMLParser import (
+        HTMLParser as compat_HTMLParser,
+        HTMLParseError as compat_HTMLParseError)
+except ImportError:  # Python 3
     from html.parser import HTMLParser as compat_HTMLParser
-except ImportError:  # Python 2
-    from HTMLParser import HTMLParser as compat_HTMLParser
-compat_html_parser_HTMLParser = compat_HTMLParser
-
-try:  # Python 2
-    from HTMLParser import HTMLParseError as compat_HTMLParseError
-except ImportError:  # Python <3.4
     try:
         from html.parser import HTMLParseError as compat_HTMLParseError
     except ImportError:  # Python >3.4
-
-        # HTMLParseError has been deprecated in Python 3.3 and removed in
+        # HTMLParseError was deprecated in Python 3.3 and removed in
         # Python 3.5. Introducing dummy exception for Python >3.5 for compatible
         # and uniform cross-version exception handling
         class compat_HTMLParseError(Exception):
             pass
+compat_html_parser_HTMLParser = compat_HTMLParser
 compat_html_parser_HTMLParseError = compat_HTMLParseError
 
 try:
-    from subprocess import DEVNULL
-    compat_subprocess_get_DEVNULL = lambda: DEVNULL
-except ImportError:
+    _DEVNULL = subprocess.DEVNULL
+    compat_subprocess_get_DEVNULL = lambda: _DEVNULL
+except AttributeError:
     compat_subprocess_get_DEVNULL = lambda: open(os.path.devnull, 'w')
 
 try:
@@ -2936,6 +2940,51 @@ else:
     compat_socket_create_connection = socket.create_connection
 
 
+try:
+    from contextlib import suppress as compat_contextlib_suppress
+except ImportError:
+    class compat_contextlib_suppress(object):
+        _exceptions = None
+
+        def __init__(self, *exceptions):
+            super(compat_contextlib_suppress, self).__init__()
+            # TODO: [Base]ExceptionGroup (3.12+)
+            self._exceptions = exceptions
+
+        def __enter__(self):
+            return self
+
+        def __exit__(self, exc_type, exc_val, exc_tb):
+            return exc_val is not None and isinstance(exc_val, self._exceptions or tuple())
+
+
+# subprocess.Popen context manager
+# avoids leaking handles if .communicate() is not called
+try:
+    _Popen = subprocess.Popen
+    # check for required context manager attributes
+    _Popen.__enter__ and _Popen.__exit__
+    compat_subprocess_Popen = _Popen
+except AttributeError:
+    # not a context manager - make one
+    from contextlib import contextmanager
+
+    @contextmanager
+    def compat_subprocess_Popen(*args, **kwargs):
+        popen = None
+        try:
+            popen = _Popen(*args, **kwargs)
+            yield popen
+        finally:
+            if popen:
+                for f in (popen.stdin, popen.stdout, popen.stderr):
+                    if f:
+                        # repeated .close() is OK, but just in case
+                        with compat_contextlib_suppress(EnvironmentError):
+                            f.close()
+                popen.wait()
+
+
 # Fix https://github.com/ytdl-org/youtube-dl/issues/4223
 # See http://bugs.python.org/issue9161 for what is broken
 def workaround_optparse_bug9161():
@@ -3256,6 +3305,7 @@ __all__ = [
     'compat_http_cookiejar_Cookie',
     'compat_http_cookies',
     'compat_http_cookies_SimpleCookie',
+    'compat_contextlib_suppress',
     'compat_ctypes_WINFUNCTYPE',
     'compat_etree_fromstring',
     'compat_filter',
@@ -3291,6 +3341,7 @@ __all__ = [
     'compat_struct_pack',
     'compat_struct_unpack',
     'compat_subprocess_get_DEVNULL',
+    'compat_subprocess_Popen',
     'compat_tokenize_tokenize',
     'compat_urllib_error',
     'compat_urllib_parse',

youtube_dl/downloader/dash.py

@@ -35,6 +35,7 @@ class DashSegmentsFD(FragmentFD):
         for frag_index, fragment in enumerate(fragments, 1):
             if frag_index <= ctx['fragment_index']:
                 continue
+            success = False
             # In DASH, the first segment contains necessary headers to
             # generate a valid MP4 file, so always abort for the first segment
             fatal = frag_index == 1 or not skip_unavailable_fragments
@@ -42,10 +43,14 @@ class DashSegmentsFD(FragmentFD):
             if not fragment_url:
                 assert fragment_base_url
                 fragment_url = urljoin(fragment_base_url, fragment['path'])
-            success = False
+            headers = info_dict.get('http_headers')
+            fragment_range = fragment.get('range')
+            if fragment_range:
+                headers = headers.copy() if headers else {}
+                headers['Range'] = 'bytes=%s' % (fragment_range,)
             for count in itertools.count():
                 try:
-                    success, frag_content = self._download_fragment(ctx, fragment_url, info_dict)
+                    success, frag_content = self._download_fragment(ctx, fragment_url, info_dict, headers)
                     if not success:
                         return False
                     self._append_fragment(ctx, frag_content)

youtube_dl/downloader/external.py

@@ -11,8 +11,14 @@ from .common import FileDownloader
 from ..compat import (
     compat_setenv,
     compat_str,
+    compat_subprocess_Popen,
 )
-from ..postprocessor.ffmpeg import FFmpegPostProcessor, EXT_TO_OUT_FORMATS
+
+try:
+    from ..postprocessor.ffmpeg import FFmpegPostProcessor, EXT_TO_OUT_FORMATS
+except ImportError:
+    FFmpegPostProcessor = None
+
 from ..utils import (
     cli_option,
     cli_valueless_option,
@@ -206,7 +212,10 @@ class WgetFD(ExternalFD):
                 retry[1] = '0'
             cmd += retry
         cmd += self._option('--bind-address', 'source_address')
-        cmd += self._option('--proxy', 'proxy')
+        proxy = self.params.get('proxy')
+        if proxy:
+            for var in ('http_proxy', 'https_proxy'):
+                cmd += ['--execute', '%s=%s' % (var, proxy)]
         cmd += self._valueless_option('--no-check-certificate', 'nocheckcertificate')
         cmd += self._configuration_args()
         cmd += ['--', info_dict['url']]
@@ -358,13 +367,14 @@ class FFmpegFD(ExternalFD):
 
     @classmethod
     def available(cls):
-        return FFmpegPostProcessor().available
+        # actual availability can only be confirmed for an instance
+        return bool(FFmpegPostProcessor)
 
     def _call_downloader(self, tmpfilename, info_dict):
-        url = info_dict['url']
-        ffpp = FFmpegPostProcessor(downloader=self)
+        # `downloader` means the parent `YoutubeDL`
+        ffpp = FFmpegPostProcessor(downloader=self.ydl)
         if not ffpp.available:
-            self.report_error('m3u8 download detected but ffmpeg or avconv could not be found. Please install one.')
+            self.report_error('ffmpeg required for download but no ffmpeg (nor avconv) executable could be found. Please install one.')
             return False
         ffpp.check_version()
 
@@ -393,6 +403,7 @@ class FFmpegFD(ExternalFD):
         # if end_time:
         #     args += ['-t', compat_str(end_time - start_time)]
 
+        url = info_dict['url']
         cookies = self.ydl.cookiejar.get_cookies_for_url(url)
         if cookies:
             args.extend(['-cookies', ''.join(
@@ -480,21 +491,25 @@ class FFmpegFD(ExternalFD):
 
         self._debug_cmd(args)
 
-        proc = subprocess.Popen(args, stdin=subprocess.PIPE, env=env)
-        try:
-            retval = proc.wait()
-        except BaseException as e:
-            # subprocess.run would send the SIGKILL signal to ffmpeg and the
-            # mp4 file couldn't be played, but if we ask ffmpeg to quit it
-            # produces a file that is playable (this is mostly useful for live
-            # streams). Note that Windows is not affected and produces playable
-            # files (see https://github.com/ytdl-org/youtube-dl/issues/8300).
-            if isinstance(e, KeyboardInterrupt) and sys.platform != 'win32':
-                process_communicate_or_kill(proc, b'q')
-            else:
-                proc.kill()
-                proc.wait()
-            raise
+        # From [1], a PIPE opened in Popen() should be closed, unless
+        # .communicate() is called. Avoid leaking any PIPEs by using Popen
+        # as a context manager (newer Python 3.x and compat)
+        # Fixes "Resource Warning" in test/test_downloader_external.py
+        # [1] https://devpress.csdn.net/python/62fde12d7e66823466192e48.html
+        with compat_subprocess_Popen(args, stdin=subprocess.PIPE, env=env) as proc:
+            try:
+                retval = proc.wait()
+            except BaseException as e:
+                # subprocess.run would send the SIGKILL signal to ffmpeg and the
+                # mp4 file couldn't be played, but if we ask ffmpeg to quit it
+                # produces a file that is playable (this is mostly useful for live
+                # streams). Note that Windows is not affected and produces playable
+                # files (see https://github.com/ytdl-org/youtube-dl/issues/8300).
+                if isinstance(e, KeyboardInterrupt) and sys.platform != 'win32':
+                    process_communicate_or_kill(proc, b'q')
+                else:
+                    proc.kill()
+                raise
         return retval
 
 

youtube_dl/extractor/caffeine.py

@@ -0,0 +1,79 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    determine_ext,
+    int_or_none,
+    merge_dicts,
+    parse_iso8601,
+    T,
+    traverse_obj,
+    txt_or_none,
+    urljoin,
+)
+
+
+class CaffeineTVIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?caffeine\.tv/[^/]+/video/(?P<id>[0-9a-f-]+)'
+    _TESTS = [{
+        'url': 'https://www.caffeine.tv/TsuSurf/video/cffc0a00-e73f-11ec-8080-80017d29f26e',
+        'info_dict': {
+            'id': 'cffc0a00-e73f-11ec-8080-80017d29f26e',
+            'ext': 'mp4',
+            'title': 'GOOOOD MORNINNNNN #highlights',
+            'timestamp': 1654702180,
+            'upload_date': '20220608',
+            'uploader': 'TsuSurf',
+            'duration': 3145,
+            'age_limit': 17,
+        },
+        'params': {
+            'format': 'bestvideo',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        json_data = self._download_json(
+            'https://api.caffeine.tv/social/public/activity/' + video_id,
+            video_id)
+        broadcast_info = traverse_obj(json_data, ('broadcast_info', T(dict))) or {}
+        title = broadcast_info['broadcast_title']
+        video_url = broadcast_info['video_url']
+
+        ext = determine_ext(video_url)
+        if ext == 'm3u8':
+            formats = self._extract_m3u8_formats(
+                video_url, video_id, 'mp4', entry_protocol='m3u8',
+                fatal=False)
+        else:
+            formats = [{'url': video_url}]
+        self._sort_formats(formats)
+
+        return merge_dicts({
+            'id': video_id,
+            'title': title,
+            'formats': formats,
+        }, traverse_obj(json_data, {
+            'uploader': ((None, 'user'), 'username'),
+        }, get_all=False), traverse_obj(json_data, {
+            'like_count': ('like_count', T(int_or_none)),
+            'view_count': ('view_count', T(int_or_none)),
+            'comment_count': ('comment_count', T(int_or_none)),
+            'tags': ('tags', Ellipsis, T(txt_or_none)),
+            'is_live': 'is_live',
+            'uploader': ('user', 'name'),
+        }), traverse_obj(broadcast_info, {
+            'duration': ('content_duration', T(int_or_none)),
+            'timestamp': ('broadcast_start_time', T(parse_iso8601)),
+            'thumbnail': ('preview_image_path', T(lambda u: urljoin(url, u))),
+            'age_limit': ('content_rating', T(lambda r: r and {
+                # assume Apple Store ratings [1]
+                # 1. https://en.wikipedia.org/wiki/Mobile_software_content_rating_system
+                'FOUR_PLUS': 0,
+                'NINE_PLUS': 9,
+                'TWELVE_PLUS': 12,
+                'SEVENTEEN_PLUS': 17,
+            }.get(r, 17))),
+        }))

youtube_dl/extractor/common.py

@@ -2,6 +2,7 @@
 from __future__ import unicode_literals
 
 import base64
+import collections
 import datetime
 import functools
 import hashlib
@@ -24,6 +25,7 @@ from ..compat import (
     compat_getpass,
     compat_integer_types,
     compat_http_client,
+    compat_kwargs,
     compat_map as map,
     compat_open as open,
     compat_os_name,
@@ -58,6 +60,7 @@ from ..utils import (
     GeoRestrictedError,
     GeoUtils,
     int_or_none,
+    join_nonempty,
     js_to_json,
     JSON_LD_RE,
     mimetype2ext,
@@ -74,6 +77,7 @@ from ..utils import (
     str_or_none,
     str_to_int,
     strip_or_none,
+    T,
     traverse_obj,
     try_get,
     unescapeHTML,
@@ -180,6 +184,8 @@ class InfoExtractor(object):
                                             fragment_base_url
                                  * "duration" (optional, int or float)
                                  * "filesize" (optional, int)
+                                 * "range" (optional, str of the form "start-end"
+                                            to use in HTTP Range header)
                     * preference Order number of this format. If this field is
                                  present and not None, the formats get sorted
                                  by this field, regardless of all other values.
@@ -596,6 +602,14 @@ class InfoExtractor(object):
         """Sets the downloader for this IE."""
         self._downloader = downloader
 
+    @property
+    def cache(self):
+        return self._downloader.cache
+
+    @property
+    def cookiejar(self):
+        return self._downloader.cookiejar
+
     def _real_initialize(self):
         """Real initialization process. Redefine in subclasses."""
         pass
@@ -942,14 +956,47 @@ class InfoExtractor(object):
             else:
                 self.report_warning(errmsg + str(ve))
 
-    def report_warning(self, msg, video_id=None):
+    def __ie_msg(self, *msg):
+        return '[{0}] {1}'.format(self.IE_NAME, ''.join(msg))
+
+    # msg, video_id=None, *args, only_once=False, **kwargs
+    def report_warning(self, msg, *args, **kwargs):
+        if len(args) > 0:
+            video_id = args[0]
+            args = args[1:]
+        else:
+            video_id = kwargs.pop('video_id', None)
         idstr = '' if video_id is None else '%s: ' % video_id
         self._downloader.report_warning(
-            '[%s] %s%s' % (self.IE_NAME, idstr, msg))
+            self.__ie_msg(idstr, msg), *args, **kwargs)
 
     def to_screen(self, msg):
         """Print msg to screen, prefixing it with '[ie_name]'"""
-        self._downloader.to_screen('[%s] %s' % (self.IE_NAME, msg))
+        self._downloader.to_screen(self.__ie_msg(msg))
+
+    def write_debug(self, msg, only_once=False, _cache=[]):
+        '''Log debug message or Print message to stderr'''
+        if not self.get_param('verbose', False):
+            return
+        message = '[debug] ' + self.__ie_msg(msg)
+        logger = self.get_param('logger')
+        if logger:
+            logger.debug(message)
+        else:
+            if only_once and hash(message) in _cache:
+                return
+            self._downloader.to_stderr(message)
+            _cache.append(hash(message))
+
+    # name, default=None, *args, **kwargs
+    def get_param(self, name, *args, **kwargs):
+        default, args = (args[0], args[1:]) if len(args) > 0 else (kwargs.pop('default', None), args)
+        if self._downloader:
+            return self._downloader.params.get(name, default, *args, **kwargs)
+        return default
+
+    def report_drm(self, video_id):
+        self.raise_no_formats('This video is DRM protected', expected=True, video_id=video_id)
 
     def report_extraction(self, id_or_name):
         """Report information extraction."""
@@ -977,6 +1024,15 @@ class InfoExtractor(object):
     def raise_geo_restricted(msg='This video is not available from your location due to geo restriction', countries=None):
         raise GeoRestrictedError(msg, countries=countries)
 
+    def raise_no_formats(self, msg, expected=False, video_id=None):
+        if expected and (
+                self.get_param('ignore_no_formats_error') or self.get_param('wait_for_video')):
+            self.report_warning(msg, video_id)
+        elif isinstance(msg, ExtractorError):
+            raise msg
+        else:
+            raise ExtractorError(msg, expected=expected, video_id=video_id)
+
     # Methods for following #608
     @staticmethod
     def url_result(url, ie=None, video_id=None, video_title=None):
@@ -1047,6 +1103,60 @@ class InfoExtractor(object):
             self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message())
             return None
 
+    def _search_json(self, start_pattern, string, name, video_id, **kwargs):
+        """Searches string for the JSON object specified by start_pattern"""
+
+        # self, start_pattern, string, name, video_id, *, end_pattern='',
+        # contains_pattern=r'{(?s:.+)}', fatal=True, default=NO_DEFAULT
+        # NB: end_pattern is only used to reduce the size of the initial match
+        end_pattern = kwargs.pop('end_pattern', '')
+        # (?:[\s\S]) simulates (?(s):.) (eg)
+        contains_pattern = kwargs.pop('contains_pattern', r'{[\s\S]+}')
+        fatal = kwargs.pop('fatal', True)
+        default = kwargs.pop('default', NO_DEFAULT)
+
+        if default is NO_DEFAULT:
+            default, has_default = {}, False
+        else:
+            fatal, has_default = False, True
+
+        json_string = self._search_regex(
+            r'(?:{0})\s*(?P<json>{1})\s*(?:{2})'.format(
+                start_pattern, contains_pattern, end_pattern),
+            string, name, group='json', fatal=fatal, default=None if has_default else NO_DEFAULT)
+        if not json_string:
+            return default
+
+        # yt-dlp has a special JSON parser that allows trailing text.
+        # Until that arrives here, the diagnostic from the exception
+        # raised by json.loads() is used to extract the wanted text.
+        # Either way, it's a problem if a transform_source() can't
+        # handle the trailing text.
+
+        # force an exception
+        kwargs['fatal'] = True
+
+        # self._downloader._format_err(name, self._downloader.Styles.EMPHASIS)
+        for _ in range(2):
+            try:
+                # return self._parse_json(json_string, video_id, ignore_extra=True, **kwargs)
+                transform_source = kwargs.pop('transform_source', None)
+                if transform_source:
+                    json_string = transform_source(json_string)
+                return self._parse_json(json_string, video_id, **compat_kwargs(kwargs))
+            except ExtractorError as e:
+                end = int_or_none(self._search_regex(r'\(char\s+(\d+)', error_to_compat_str(e), 'end', default=None))
+                if end is not None:
+                    json_string = json_string[:end]
+                    continue
+                msg = 'Unable to extract {0} - Failed to parse JSON'.format(name)
+                if fatal:
+                    raise ExtractorError(msg, cause=e.cause, video_id=video_id)
+                elif not has_default:
+                    self.report_warning(
+                        '{0}: {1}'.format(msg, error_to_compat_str(e)), video_id=video_id)
+            return default
+
     def _html_search_regex(self, pattern, string, name, default=NO_DEFAULT, fatal=True, flags=0, group=None):
         """
         Like _search_regex, but strips HTML tags and unescapes entities.
@@ -1701,6 +1811,12 @@ class InfoExtractor(object):
             'format_note': 'Quality selection URL',
         }
 
+    def _report_ignoring_subs(self, name):
+        self.report_warning(bug_reports_message(
+            'Ignoring subtitle tracks found in the {0} manifest; '
+            'if any subtitle tracks are missing,'.format(name)
+        ), only_once=True)
+
     def _extract_m3u8_formats(self, m3u8_url, video_id, ext=None,
                               entry_protocol='m3u8', preference=None,
                               m3u8_id=None, note=None, errnote=None,
@@ -2141,23 +2257,46 @@ class InfoExtractor(object):
             })
         return entries
 
-    def _extract_mpd_formats(self, mpd_url, video_id, mpd_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
+    def _extract_mpd_formats(self, *args, **kwargs):
+        fmts, subs = self._extract_mpd_formats_and_subtitles(*args, **kwargs)
+        if subs:
+            self._report_ignoring_subs('DASH')
+        return fmts
+
+    def _extract_mpd_formats_and_subtitles(
+            self, mpd_url, video_id, mpd_id=None, note=None, errnote=None,
+            fatal=True, data=None, headers=None, query=None):
+
+        # TODO: or not? param not yet implemented
+        if self.get_param('ignore_no_formats_error'):
+            fatal = False
+
         res = self._download_xml_handle(
             mpd_url, video_id,
-            note=note or 'Downloading MPD manifest',
-            errnote=errnote or 'Failed to download MPD manifest',
-            fatal=fatal, data=data, headers=headers, query=query)
+            note='Downloading MPD manifest' if note is None else note,
+            errnote='Failed to download MPD manifest' if errnote is None else errnote,
+            fatal=fatal, data=data, headers=headers or {}, query=query or {})
         if res is False:
-            return []
+            return [], {}
         mpd_doc, urlh = res
         if mpd_doc is None:
-            return []
-        mpd_base_url = base_url(urlh.geturl())
+            return [], {}
 
-        return self._parse_mpd_formats(
+        # We could have been redirected to a new url when we retrieved our mpd file.
+        mpd_url = urlh.geturl()
+        mpd_base_url = base_url(mpd_url)
+
+        return self._parse_mpd_formats_and_subtitles(
             mpd_doc, mpd_id, mpd_base_url, mpd_url)
 
-    def _parse_mpd_formats(self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None):
+    def _parse_mpd_formats(self, *args, **kwargs):
+        fmts, subs = self._parse_mpd_formats_and_subtitles(*args, **kwargs)
+        if subs:
+            self._report_ignoring_subs('DASH')
+        return fmts
+
+    def _parse_mpd_formats_and_subtitles(
+            self, mpd_doc, mpd_id=None, mpd_base_url='', mpd_url=None):
         """
         Parse formats from MPD manifest.
         References:
@@ -2165,8 +2304,10 @@ class InfoExtractor(object):
             http://standards.iso.org/ittf/PubliclyAvailableStandards/c065274_ISO_IEC_23009-1_2014.zip
          2. https://en.wikipedia.org/wiki/Dynamic_Adaptive_Streaming_over_HTTP
         """
-        if mpd_doc.get('type') == 'dynamic':
-            return []
+        # TODO: param not yet implemented: default like previous yt-dl logic
+        if not self.get_param('dynamic_mpd', False):
+            if mpd_doc.get('type') == 'dynamic':
+                return [], {}
 
         namespace = self._search_regex(r'(?i)^{([^}]+)?}MPD$', mpd_doc.tag, 'namespace', default=None)
 
@@ -2176,8 +2317,24 @@ class InfoExtractor(object):
         def is_drm_protected(element):
             return element.find(_add_ns('ContentProtection')) is not None
 
+        from ..utils import YoutubeDLHandler
+        fix_path = YoutubeDLHandler._fix_path
+
+        def resolve_base_url(element, parent_base_url=None):
+            # TODO: use native XML traversal when ready
+            b_url = traverse_obj(element, (
+                T(lambda e: e.find(_add_ns('BaseURL')).text)))
+            if parent_base_url and b_url:
+                if not parent_base_url[-1] in ('/', ':'):
+                    parent_base_url += '/'
+                b_url = compat_urlparse.urljoin(parent_base_url, b_url)
+            if b_url:
+                b_url = fix_path(b_url)
+            return b_url or parent_base_url
+
         def extract_multisegment_info(element, ms_parent_info):
             ms_info = ms_parent_info.copy()
+            base_url = ms_info['base_url'] = resolve_base_url(element, ms_info.get('base_url'))
 
             # As per [1, 5.3.9.2.2] SegmentList and SegmentTemplate share some
             # common attributes and elements.  We will only extract relevant
@@ -2211,15 +2368,27 @@ class InfoExtractor(object):
             def extract_Initialization(source):
                 initialization = source.find(_add_ns('Initialization'))
                 if initialization is not None:
-                    ms_info['initialization_url'] = initialization.attrib['sourceURL']
+                    ms_info['initialization_url'] = initialization.get('sourceURL') or base_url
+                    initialization_url_range = initialization.get('range')
+                    if initialization_url_range:
+                        ms_info['initialization_url_range'] = initialization_url_range
 
             segment_list = element.find(_add_ns('SegmentList'))
             if segment_list is not None:
                 extract_common(segment_list)
                 extract_Initialization(segment_list)
                 segment_urls_e = segment_list.findall(_add_ns('SegmentURL'))
-                if segment_urls_e:
-                    ms_info['segment_urls'] = [segment.attrib['media'] for segment in segment_urls_e]
+                segment_urls = traverse_obj(segment_urls_e, (
+                    Ellipsis, T(lambda e: e.attrib), 'media'))
+                if segment_urls:
+                    ms_info['segment_urls'] = segment_urls
+                segment_urls_range = traverse_obj(segment_urls_e, (
+                    Ellipsis, T(lambda e: e.attrib), 'mediaRange',
+                    T(lambda r: re.findall(r'^\d+-\d+$', r)), 0))
+                if segment_urls_range:
+                    ms_info['segment_urls_range'] = segment_urls_range
+                    if not segment_urls:
+                        ms_info['segment_urls'] = [base_url for _ in segment_urls_range]
             else:
                 segment_template = element.find(_add_ns('SegmentTemplate'))
                 if segment_template is not None:
@@ -2235,17 +2404,20 @@ class InfoExtractor(object):
             return ms_info
 
         mpd_duration = parse_duration(mpd_doc.get('mediaPresentationDuration'))
-        formats = []
+        formats, subtitles = [], {}
+        stream_numbers = collections.defaultdict(int)
+        mpd_base_url = resolve_base_url(mpd_doc, mpd_base_url or mpd_url)
         for period in mpd_doc.findall(_add_ns('Period')):
             period_duration = parse_duration(period.get('duration')) or mpd_duration
             period_ms_info = extract_multisegment_info(period, {
                 'start_number': 1,
                 'timescale': 1,
+                'base_url': mpd_base_url,
             })
             for adaptation_set in period.findall(_add_ns('AdaptationSet')):
                 if is_drm_protected(adaptation_set):
                     continue
-                adaption_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
+                adaptation_set_ms_info = extract_multisegment_info(adaptation_set, period_ms_info)
                 for representation in adaptation_set.findall(_add_ns('Representation')):
                     if is_drm_protected(representation):
                         continue
@@ -2253,27 +2425,35 @@ class InfoExtractor(object):
                     representation_attrib.update(representation.attrib)
                     # According to [1, 5.3.7.2, Table 9, page 41], @mimeType is mandatory
                     mime_type = representation_attrib['mimeType']
-                    content_type = mime_type.split('/')[0]
-                    if content_type == 'text':
-                        # TODO implement WebVTT downloading
-                        pass
-                    elif content_type in ('video', 'audio'):
-                        base_url = ''
-                        for element in (representation, adaptation_set, period, mpd_doc):
-                            base_url_e = element.find(_add_ns('BaseURL'))
-                            if base_url_e is not None:
-                                base_url = base_url_e.text + base_url
-                                if re.match(r'^https?://', base_url):
-                                    break
-                        if mpd_base_url and not re.match(r'^https?://', base_url):
-                            if not mpd_base_url.endswith('/') and not base_url.startswith('/'):
-                                mpd_base_url += '/'
-                            base_url = mpd_base_url + base_url
-                        representation_id = representation_attrib.get('id')
-                        lang = representation_attrib.get('lang')
-                        url_el = representation.find(_add_ns('BaseURL'))
-                        filesize = int_or_none(url_el.attrib.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
-                        bandwidth = int_or_none(representation_attrib.get('bandwidth'))
+                    content_type = representation_attrib.get('contentType') or mime_type.split('/')[0]
+                    codec_str = representation_attrib.get('codecs', '')
+                    # Some kind of binary subtitle found in some youtube livestreams
+                    if mime_type == 'application/x-rawcc':
+                        codecs = {'scodec': codec_str}
+                    else:
+                        codecs = parse_codecs(codec_str)
+                    if content_type not in ('video', 'audio', 'text'):
+                        if mime_type == 'image/jpeg':
+                            content_type = mime_type
+                        elif codecs.get('vcodec', 'none') != 'none':
+                            content_type = 'video'
+                        elif codecs.get('acodec', 'none') != 'none':
+                            content_type = 'audio'
+                        elif codecs.get('scodec', 'none') != 'none':
+                            content_type = 'text'
+                        elif mimetype2ext(mime_type) in ('tt', 'dfxp', 'ttml', 'xml', 'json'):
+                            content_type = 'text'
+                        else:
+                            self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
+                            continue
+
+                    representation_id = representation_attrib.get('id')
+                    lang = representation_attrib.get('lang')
+                    url_el = representation.find(_add_ns('BaseURL'))
+                    filesize = int_or_none(url_el.get('{http://youtube.com/yt/2012/10/10}contentLength') if url_el is not None else None)
+                    bandwidth = int_or_none(representation_attrib.get('bandwidth'))
+                    format_id = join_nonempty(representation_id or content_type, mpd_id)
+                    if content_type in ('video', 'audio'):
                         f = {
                             'format_id': '%s-%s' % (mpd_id, representation_id) if mpd_id else representation_id,
                             'manifest_url': mpd_url,
@@ -2288,104 +2468,130 @@ class InfoExtractor(object):
                             'filesize': filesize,
                             'container': mimetype2ext(mime_type) + '_dash',
                         }
-                        f.update(parse_codecs(representation_attrib.get('codecs')))
-                        representation_ms_info = extract_multisegment_info(representation, adaption_set_ms_info)
+                        f.update(codecs)
+                    elif content_type == 'text':
+                        f = {
+                            'ext': mimetype2ext(mime_type),
+                            'manifest_url': mpd_url,
+                            'filesize': filesize,
+                        }
+                    elif content_type == 'image/jpeg':
+                        # See test case in VikiIE
+                        # https://www.viki.com/videos/1175236v-choosing-spouse-by-lottery-episode-1
+                        f = {
+                            'format_id': format_id,
+                            'ext': 'mhtml',
+                            'manifest_url': mpd_url,
+                            'format_note': 'DASH storyboards (jpeg)',
+                            'acodec': 'none',
+                            'vcodec': 'none',
+                        }
+                    if is_drm_protected(adaptation_set) or is_drm_protected(representation):
+                        f['has_drm'] = True
+                    representation_ms_info = extract_multisegment_info(representation, adaptation_set_ms_info)
 
-                        def prepare_template(template_name, identifiers):
-                            tmpl = representation_ms_info[template_name]
-                            # First of, % characters outside $...$ templates
-                            # must be escaped by doubling for proper processing
-                            # by % operator string formatting used further (see
-                            # https://github.com/ytdl-org/youtube-dl/issues/16867).
-                            t = ''
-                            in_template = False
-                            for c in tmpl:
+                    def prepare_template(template_name, identifiers):
+                        tmpl = representation_ms_info[template_name]
+                        # First of, % characters outside $...$ templates
+                        # must be escaped by doubling for proper processing
+                        # by % operator string formatting used further (see
+                        # https://github.com/ytdl-org/youtube-dl/issues/16867).
+                        t = ''
+                        in_template = False
+                        for c in tmpl:
+                            t += c
+                            if c == '$':
+                                in_template = not in_template
+                            elif c == '%' and not in_template:
                                 t += c
-                                if c == '$':
-                                    in_template = not in_template
-                                elif c == '%' and not in_template:
-                                    t += c
-                            # Next, $...$ templates are translated to their
-                            # %(...) counterparts to be used with % operator
-                            t = t.replace('$RepresentationID$', representation_id)
-                            t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t)
-                            t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t)
-                            t.replace('$$', '$')
-                            return t
+                        # Next, $...$ templates are translated to their
+                        # %(...) counterparts to be used with % operator
+                        t = t.replace('$RepresentationID$', representation_id)
+                        t = re.sub(r'\$(%s)\$' % '|'.join(identifiers), r'%(\1)d', t)
+                        t = re.sub(r'\$(%s)%%([^$]+)\$' % '|'.join(identifiers), r'%(\1)\2', t)
+                        t.replace('$$', '$')
+                        return t
 
-                        # @initialization is a regular template like @media one
-                        # so it should be handled just the same way (see
-                        # https://github.com/ytdl-org/youtube-dl/issues/11605)
-                        if 'initialization' in representation_ms_info:
-                            initialization_template = prepare_template(
-                                'initialization',
-                                # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
-                                # $Time$ shall not be included for @initialization thus
-                                # only $Bandwidth$ remains
-                                ('Bandwidth', ))
-                            representation_ms_info['initialization_url'] = initialization_template % {
-                                'Bandwidth': bandwidth,
-                            }
+                    # @initialization is a regular template like @media one
+                    # so it should be handled just the same way (see
+                    # https://github.com/ytdl-org/youtube-dl/issues/11605)
+                    if 'initialization' in representation_ms_info:
+                        initialization_template = prepare_template(
+                            'initialization',
+                            # As per [1, 5.3.9.4.2, Table 15, page 54] $Number$ and
+                            # $Time$ shall not be included for @initialization thus
+                            # only $Bandwidth$ remains
+                            ('Bandwidth', ))
+                        representation_ms_info['initialization_url'] = initialization_template % {
+                            'Bandwidth': bandwidth,
+                        }
 
-                        def location_key(location):
-                            return 'url' if re.match(r'^https?://', location) else 'path'
+                    def location_key(location):
+                        return 'url' if re.match(r'^https?://', location) else 'path'
 
-                        if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info:
+                    def calc_segment_duration():
+                        return float_or_none(
+                            representation_ms_info['segment_duration'],
+                            representation_ms_info['timescale']) if 'segment_duration' in representation_ms_info else None
 
-                            media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time'))
-                            media_location_key = location_key(media_template)
+                    if 'segment_urls' not in representation_ms_info and 'media' in representation_ms_info:
 
-                            # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
-                            # can't be used at the same time
-                            if '%(Number' in media_template and 's' not in representation_ms_info:
-                                segment_duration = None
-                                if 'total_number' not in representation_ms_info and 'segment_duration' in representation_ms_info:
-                                    segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale'])
-                                    representation_ms_info['total_number'] = int(math.ceil(float(period_duration) / segment_duration))
-                                representation_ms_info['fragments'] = [{
-                                    media_location_key: media_template % {
-                                        'Number': segment_number,
-                                        'Bandwidth': bandwidth,
-                                    },
-                                    'duration': segment_duration,
-                                } for segment_number in range(
-                                    representation_ms_info['start_number'],
-                                    representation_ms_info['total_number'] + representation_ms_info['start_number'])]
-                            else:
-                                # $Number*$ or $Time$ in media template with S list available
-                                # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
-                                # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
-                                representation_ms_info['fragments'] = []
-                                segment_time = 0
-                                segment_d = None
-                                segment_number = representation_ms_info['start_number']
+                        media_template = prepare_template('media', ('Number', 'Bandwidth', 'Time'))
+                        media_location_key = location_key(media_template)
 
-                                def add_segment_url():
-                                    segment_url = media_template % {
-                                        'Time': segment_time,
-                                        'Bandwidth': bandwidth,
-                                        'Number': segment_number,
-                                    }
-                                    representation_ms_info['fragments'].append({
-                                        media_location_key: segment_url,
-                                        'duration': float_or_none(segment_d, representation_ms_info['timescale']),
-                                    })
+                        # As per [1, 5.3.9.4.4, Table 16, page 55] $Number$ and $Time$
+                        # can't be used at the same time
+                        if '%(Number' in media_template and 's' not in representation_ms_info:
+                            segment_duration = None
+                            if 'total_number' not in representation_ms_info and 'segment_duration' in representation_ms_info:
+                                segment_duration = float_or_none(representation_ms_info['segment_duration'], representation_ms_info['timescale'])
+                                representation_ms_info['total_number'] = int(math.ceil(
+                                    float_or_none(period_duration, segment_duration, default=0)))
+                            representation_ms_info['fragments'] = [{
+                                media_location_key: media_template % {
+                                    'Number': segment_number,
+                                    'Bandwidth': bandwidth,
+                                },
+                                'duration': segment_duration,
+                            } for segment_number in range(
+                                representation_ms_info['start_number'],
+                                representation_ms_info['total_number'] + representation_ms_info['start_number'])]
+                        else:
+                            # $Number*$ or $Time$ in media template with S list available
+                            # Example $Number*$: http://www.svtplay.se/klipp/9023742/stopptid-om-bjorn-borg
+                            # Example $Time$: https://play.arkena.com/embed/avp/v2/player/media/b41dda37-d8e7-4d3f-b1b5-9a9db578bdfe/1/129411
+                            representation_ms_info['fragments'] = []
+                            segment_time = 0
+                            segment_d = None
+                            segment_number = representation_ms_info['start_number']
 
-                                for num, s in enumerate(representation_ms_info['s']):
-                                    segment_time = s.get('t') or segment_time
-                                    segment_d = s['d']
+                            def add_segment_url():
+                                segment_url = media_template % {
+                                    'Time': segment_time,
+                                    'Bandwidth': bandwidth,
+                                    'Number': segment_number,
+                                }
+                                representation_ms_info['fragments'].append({
+                                    media_location_key: segment_url,
+                                    'duration': float_or_none(segment_d, representation_ms_info['timescale']),
+                                })
+
+                            for num, s in enumerate(representation_ms_info['s']):
+                                segment_time = s.get('t') or segment_time
+                                segment_d = s['d']
+                                add_segment_url()
+                                segment_number += 1
+                                for r in range(s.get('r', 0)):
+                                    segment_time += segment_d
                                     add_segment_url()
                                     segment_number += 1
-                                    for r in range(s.get('r', 0)):
-                                        segment_time += segment_d
-                                        add_segment_url()
-                                        segment_number += 1
-                                    segment_time += segment_d
-                        elif 'segment_urls' in representation_ms_info and 's' in representation_ms_info:
+                                segment_time += segment_d
+                    elif 'segment_urls' in representation_ms_info:
+                        fragments = []
+                        if 's' in representation_ms_info:
                             # No media template
                             # Example: https://www.youtube.com/watch?v=iXZV5uAYMJI
                             # or any YouTube dashsegments video
-                            fragments = []
                             segment_index = 0
                             timescale = representation_ms_info['timescale']
                             for s in representation_ms_info['s']:
@@ -2397,48 +2603,78 @@ class InfoExtractor(object):
                                         'duration': duration,
                                     })
                                     segment_index += 1
-                            representation_ms_info['fragments'] = fragments
-                        elif 'segment_urls' in representation_ms_info:
+                        elif 'segment_urls_range' in representation_ms_info:
+                            # Segment URLs with mediaRange
+                            # Example: https://kinescope.io/200615537/master.mpd
+                            # https://github.com/ytdl-org/youtube-dl/issues/30235
+                            # or any mpd generated with Bento4 `mp4dash --no-split --use-segment-list`
+                            segment_duration = calc_segment_duration()
+                            for segment_url, segment_url_range in zip(
+                                    representation_ms_info['segment_urls'], representation_ms_info['segment_urls_range']):
+                                fragments.append({
+                                    location_key(segment_url): segment_url,
+                                    'range': segment_url_range,
+                                    'duration': segment_duration,
+                                })
+                        else:
                             # Segment URLs with no SegmentTimeline
                             # Example: https://www.seznam.cz/zpravy/clanek/cesko-zasahne-vitr-o-sile-vichrice-muze-byt-i-zivotu-nebezpecny-39091
                             # https://github.com/ytdl-org/youtube-dl/pull/14844
-                            fragments = []
-                            segment_duration = float_or_none(
-                                representation_ms_info['segment_duration'],
-                                representation_ms_info['timescale']) if 'segment_duration' in representation_ms_info else None
+                            segment_duration = calc_segment_duration()
                             for segment_url in representation_ms_info['segment_urls']:
-                                fragment = {
+                                fragments.append({
                                     location_key(segment_url): segment_url,
-                                }
-                                if segment_duration:
-                                    fragment['duration'] = segment_duration
-                                fragments.append(fragment)
-                            representation_ms_info['fragments'] = fragments
-                        # If there is a fragments key available then we correctly recognized fragmented media.
-                        # Otherwise we will assume unfragmented media with direct access. Technically, such
-                        # assumption is not necessarily correct since we may simply have no support for
-                        # some forms of fragmented media renditions yet, but for now we'll use this fallback.
-                        if 'fragments' in representation_ms_info:
-                            f.update({
-                                # NB: mpd_url may be empty when MPD manifest is parsed from a string
-                                'url': mpd_url or base_url,
-                                'fragment_base_url': base_url,
-                                'fragments': [],
-                                'protocol': 'http_dash_segments',
+                                    'duration': segment_duration,
+                                })
+                        representation_ms_info['fragments'] = fragments
+
+                    # If there is a fragments key available then we correctly recognized fragmented media.
+                    # Otherwise we will assume unfragmented media with direct access. Technically, such
+                    # assumption is not necessarily correct since we may simply have no support for
+                    # some forms of fragmented media renditions yet, but for now we'll use this fallback.
+                    if 'fragments' in representation_ms_info:
+                        base_url = representation_ms_info['base_url']
+                        f.update({
+                            # NB: mpd_url may be empty when MPD manifest is parsed from a string
+                            'url': mpd_url or base_url,
+                            'fragment_base_url': base_url,
+                            'fragments': [],
+                            'protocol': 'http_dash_segments',
+                        })
+                        if 'initialization_url' in representation_ms_info and 'initialization_url_range' in representation_ms_info:
+                            # Initialization URL with range (accompanied by Segment URLs with mediaRange above)
+                            # https://github.com/ytdl-org/youtube-dl/issues/30235
+                            initialization_url = representation_ms_info['initialization_url']
+                            f['fragments'].append({
+                                location_key(initialization_url): initialization_url,
+                                'range': representation_ms_info['initialization_url_range'],
                             })
-                            if 'initialization_url' in representation_ms_info:
-                                initialization_url = representation_ms_info['initialization_url']
-                                if not f.get('url'):
-                                    f['url'] = initialization_url
-                                f['fragments'].append({location_key(initialization_url): initialization_url})
-                            f['fragments'].extend(representation_ms_info['fragments'])
-                        else:
-                            # Assuming direct URL to unfragmented media.
-                            f['url'] = base_url
-                        formats.append(f)
+                        elif 'initialization_url' in representation_ms_info:
+                            initialization_url = representation_ms_info['initialization_url']
+                            if not f.get('url'):
+                                f['url'] = initialization_url
+                            f['fragments'].append({location_key(initialization_url): initialization_url})
+                        elif 'initialization_url_range' in representation_ms_info:
+                            # no Initialization URL but range (accompanied by no Segment URLs but mediaRange above)
+                            # https://github.com/ytdl-org/youtube-dl/issues/27575
+                            f['fragments'].append({
+                                location_key(base_url): base_url,
+                                'range': representation_ms_info['initialization_url_range'],
+                            })
+                        f['fragments'].extend(representation_ms_info['fragments'])
+                        if not period_duration:
+                            period_duration = sum(traverse_obj(representation_ms_info, (
+                                'fragments', Ellipsis, 'duration', T(float_or_none))))
                     else:
-                        self.report_warning('Unknown MIME type %s in DASH manifest' % mime_type)
-        return formats
+                        # Assuming direct URL to unfragmented media.
+                        f['url'] = representation_ms_info['base_url']
+                    if content_type in ('video', 'audio', 'image/jpeg'):
+                        f['manifest_stream_number'] = stream_numbers[f['url']]
+                        stream_numbers[f['url']] += 1
+                        formats.append(f)
+                    elif content_type == 'text':
+                        subtitles.setdefault(lang or 'und', []).append(f)
+        return formats, subtitles
 
     def _extract_ism_formats(self, ism_url, video_id, ism_id=None, note=None, errnote=None, fatal=True, data=None, headers={}, query={}):
         res = self._download_xml_handle(
@@ -2785,25 +3021,22 @@ class InfoExtractor(object):
         return formats
 
     def _find_jwplayer_data(self, webpage, video_id=None, transform_source=js_to_json):
-        mobj = re.search(
-            r'''(?s)jwplayer\s*\(\s*(?P<q>'|")(?!(?P=q)).+(?P=q)\s*\)(?!</script>).*?\.\s*setup\s*\(\s*(?P<options>(?:\([^)]*\)|[^)])+)\s*\)''',
-            webpage)
-        if mobj:
-            try:
-                jwplayer_data = self._parse_json(mobj.group('options'),
-                                                 video_id=video_id,
-                                                 transform_source=transform_source)
-            except ExtractorError:
-                pass
-            else:
-                if isinstance(jwplayer_data, dict):
-                    return jwplayer_data
+        return self._search_json(
+            r'''(?<!-)\bjwplayer\s*\(\s*(?P<q>'|")(?!(?P=q)).+(?P=q)\s*\)(?:(?!</script>).)*?\.\s*(?:setup\s*\(|(?P<load>load)\s*\(\s*\[)''',
+            webpage, 'JWPlayer data', video_id,
+            # must be a {...} or sequence, ending
+            contains_pattern=r'\{[\s\S]*}(?(load)(?:\s*,\s*\{[\s\S]*})*)', end_pattern=r'(?(load)\]|\))',
+            transform_source=transform_source, default=None)
 
     def _extract_jwplayer_data(self, webpage, video_id, *args, **kwargs):
-        jwplayer_data = self._find_jwplayer_data(
-            webpage, video_id, transform_source=js_to_json)
-        return self._parse_jwplayer_data(
-            jwplayer_data, video_id, *args, **kwargs)
+
+        # allow passing `transform_source` through to _find_jwplayer_data()
+        transform_source = kwargs.pop('transform_source', None)
+        kwfind = compat_kwargs({'transform_source': transform_source}) if transform_source else {}
+
+        jwplayer_data = self._find_jwplayer_data(webpage, video_id, **kwfind)
+
+        return self._parse_jwplayer_data(jwplayer_data, video_id, *args, **kwargs)
 
     def _parse_jwplayer_data(self, jwplayer_data, video_id=None, require_title=True,
                              m3u8_id=None, mpd_id=None, rtmp_params=None, base_url=None):
@@ -2837,22 +3070,14 @@ class InfoExtractor(object):
                 mpd_id=mpd_id, rtmp_params=rtmp_params, base_url=base_url)
 
             subtitles = {}
-            tracks = video_data.get('tracks')
-            if tracks and isinstance(tracks, list):
-                for track in tracks:
-                    if not isinstance(track, dict):
-                        continue
-                    track_kind = track.get('kind')
-                    if not track_kind or not isinstance(track_kind, compat_str):
-                        continue
-                    if track_kind.lower() not in ('captions', 'subtitles'):
-                        continue
-                    track_url = urljoin(base_url, track.get('file'))
-                    if not track_url:
-                        continue
-                    subtitles.setdefault(track.get('label') or 'en', []).append({
-                        'url': self._proto_relative_url(track_url)
-                    })
+            for track in traverse_obj(video_data, (
+                    'tracks', lambda _, t: t.get('kind').lower() in ('captions', 'subtitles'))):
+                track_url = urljoin(base_url, track.get('file'))
+                if not track_url:
+                    continue
+                subtitles.setdefault(track.get('label') or 'en', []).append({
+                    'url': self._proto_relative_url(track_url)
+                })
 
             entry = {
                 'id': this_video_id,

youtube_dl/extractor/epidemicsound.py

@@ -0,0 +1,101 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    float_or_none,
+    T,
+    traverse_obj,
+    txt_or_none,
+    unified_timestamp,
+    url_or_none,
+)
+
+
+class EpidemicSoundIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?epidemicsound\.com/track/(?P<id>[0-9a-zA-Z]+)'
+    _TESTS = [{
+        'url': 'https://www.epidemicsound.com/track/yFfQVRpSPz/',
+        'md5': 'd98ff2ddb49e8acab9716541cbc9dfac',
+        'info_dict': {
+            'id': '45014',
+            'display_id': 'yFfQVRpSPz',
+            'ext': 'mp3',
+            'tags': ['foley', 'door', 'knock', 'glass', 'window', 'glass door knock'],
+            'title': 'Door Knock Door 1',
+            'duration': 1,
+            'thumbnail': 'https://cdn.epidemicsound.com/curation-assets/commercial-release-cover-images/default-sfx/3000x3000.jpg',
+            'timestamp': 1415320353,
+            'upload_date': '20141107',
+            'age_limit': None,
+            # check that the "best" format was found, since test file MD5 doesn't
+            # distinguish the formats
+            'format': 'full',
+        },
+    }, {
+        'url': 'https://www.epidemicsound.com/track/mj8GTTwsZd/',
+        'md5': 'c82b745890f9baf18dc2f8d568ee3830',
+        'info_dict': {
+            'id': '148700',
+            'display_id': 'mj8GTTwsZd',
+            'ext': 'mp3',
+            'tags': ['liquid drum n bass', 'energetic'],
+            'title': 'Noplace',
+            'duration': 237,
+            'thumbnail': 'https://cdn.epidemicsound.com/curation-assets/commercial-release-cover-images/11138/3000x3000.jpg',
+            'timestamp': 1694426482,
+            'release_timestamp': 1700535606,
+            'upload_date': '20230911',
+            'age_limit': None,
+            'format': 'full',
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        json_data = self._download_json('https://www.epidemicsound.com/json/track/' + video_id, video_id)
+
+        def fmt_or_none(f):
+            if not f.get('format'):
+                f['format'] = f.get('format_id')
+            elif not f.get('format_id'):
+                f['format_id'] = f['format']
+            if not (f['url'] and f['format']):
+                return
+            if f.get('format_note'):
+                f['format_note'] = 'track ID ' + f['format_note']
+            f['preference'] = -1 if f['format'] == 'full' else -2
+            return f
+
+        formats = traverse_obj(json_data, (
+            'stems', T(dict.items), Ellipsis, {
+                'format': (0, T(txt_or_none)),
+                'format_note': (1, 's3TrackId', T(txt_or_none)),
+                'format_id': (1, 'stemType', T(txt_or_none)),
+                'url': (1, 'lqMp3Url', T(url_or_none)),
+            }, T(fmt_or_none)))
+
+        self._sort_formats(formats)
+
+        info = traverse_obj(json_data, {
+            'id': ('id', T(txt_or_none)),
+            'tags': ('metadataTags', Ellipsis, T(txt_or_none)),
+            'title': ('title', T(txt_or_none)),
+            'duration': ('length', T(float_or_none)),
+            'timestamp': ('added', T(unified_timestamp)),
+            'thumbnail': (('imageUrl', 'cover'), T(url_or_none)),
+            'age_limit': ('isExplicit', T(lambda b: 18 if b else None)),
+            'release_timestamp': ('releaseDate', T(unified_timestamp)),
+        }, get_all=False)
+
+        info.update(traverse_obj(json_data, {
+            'categories': ('genres', Ellipsis, 'tag', T(txt_or_none)),
+            'tags': ('metadataTags', Ellipsis, T(txt_or_none)),
+        }))
+
+        info.update({
+            'display_id': video_id,
+            'formats': formats,
+        })
+
+        return info

youtube_dl/extractor/extractors.py

@@ -159,6 +159,7 @@ from .businessinsider import BusinessInsiderIE
 from .buzzfeed import BuzzFeedIE
 from .byutv import BYUtvIE
 from .c56 import C56IE
+from .caffeine import CaffeineTVIE
 from .callin import CallinIE
 from .camdemy import (
     CamdemyIE,
@@ -357,6 +358,7 @@ from .ellentube import (
 from .elpais import ElPaisIE
 from .embedly import EmbedlyIE
 from .engadget import EngadgetIE
+from .epidemicsound import EpidemicSoundIE
 from .eporner import EpornerIE
 from .eroprofile import EroProfileIE
 from .escapist import EscapistIE
@@ -381,7 +383,6 @@ from .fc2 import (
     FC2EmbedIE,
 )
 from .fczenit import FczenitIE
-from .filemoon import FileMoonIE
 from .fifa import FifaIE
 from .filmon import (
     FilmOnIE,
@@ -442,6 +443,7 @@ from .gamespot import GameSpotIE
 from .gamestar import GameStarIE
 from .gaskrank import GaskrankIE
 from .gazeta import GazetaIE
+from .gbnews import GBNewsIE
 from .gdcvault import GDCVaultIE
 from .gedidigital import GediDigitalIE
 from .generic import GenericIE

youtube_dl/extractor/filemoon.py

@@ -1,43 +0,0 @@
-# coding: utf-8
-from __future__ import unicode_literals
-
-import re
-
-from .common import InfoExtractor
-from ..utils import (
-    decode_packed_codes,
-    js_to_json,
-)
-
-
-class FileMoonIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?filemoon\.sx/./(?P<id>\w+)'
-    _TEST = {
-        'url': 'https://filemoon.sx/e/dw40rxrzruqz',
-        'md5': '5a713742f57ac4aef29b74733e8dda01',
-        'info_dict': {
-            'id': 'dw40rxrzruqz',
-            'title': 'dw40rxrzruqz',
-            'ext': 'mp4'
-        }
-    }
-
-    def _real_extract(self, url):
-        video_id = self._match_id(url)
-
-        webpage = self._download_webpage(url, video_id)
-        matches = re.findall(r'(?s)(eval.*?)</script>', webpage)
-        packed = matches[-1]
-        unpacked = decode_packed_codes(packed)
-        jwplayer_sources = self._parse_json(
-            self._search_regex(
-                r'(?s)player\s*\.\s*setup\s*\(\s*\{\s*sources\s*:\s*(.*?])', unpacked, 'jwplayer sources'),
-            video_id, transform_source=js_to_json)
-
-        formats = self._parse_jwplayer_formats(jwplayer_sources, video_id)
-
-        return {
-            'id': video_id,
-            'title': self._generic_title(url) or video_id,
-            'formats': formats
-        }

youtube_dl/extractor/gbnews.py

@@ -0,0 +1,139 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+from .common import InfoExtractor
+from ..utils import (
+    extract_attributes,
+    ExtractorError,
+    T,
+    traverse_obj,
+    txt_or_none,
+    url_or_none,
+)
+
+
+class GBNewsIE(InfoExtractor):
+    IE_DESC = 'GB News clips, features and live stream'
+
+    # \w+ is normally shows or news, but apparently any word redirects to the correct URL
+    _VALID_URL = r'https?://(?:www\.)?gbnews\.(?:uk|com)/(?:\w+/)?(?P<id>[^#?]+)'
+
+    _PLATFORM = 'safari'
+    _SSMP_URL = 'https://mm-v2.simplestream.com/ssmp/api.php'
+    _TESTS = [{
+        'url': 'https://www.gbnews.uk/shows/andrew-neils-message-to-companies-choosing-to-boycott-gb-news/106889',
+        'info_dict': {
+            'id': '106889',
+            'ext': 'mp4',
+            'title': "Andrew Neil's message to companies choosing to boycott GB News",
+            'description': 'md5:b281f5d22fd6d5eda64a4e3ba771b351',
+        },
+        'skip': '404 not found',
+    }, {
+        'url': 'https://www.gbnews.com/news/bbc-claudine-gay-harvard-university-antisemitism-row',
+        'info_dict': {
+            'id': '52264136',
+            'display_id': 'bbc-claudine-gay-harvard-university-antisemitism-row',
+            'ext': 'mp4',
+            'title': 'BBC deletes post after furious backlash over headline downplaying antisemitism',
+            'description': 'The post was criticised by former employers of the broadcaster',
+        },
+    }, {
+        'url': 'https://www.gbnews.uk/watchlive',
+        'info_dict': {
+            'id': '1069',
+            'display_id': 'watchlive',
+            'ext': 'mp4',
+            'title': 'GB News Live',
+            'is_live': True,
+        },
+        'params': {
+            'skip_download': 'm3u8',
+        },
+    }]
+
+    def _real_extract(self, url):
+        display_id = self._match_id(url).split('/')[-1]
+
+        webpage = self._download_webpage(url, display_id)
+        # extraction based on https://github.com/ytdl-org/youtube-dl/issues/29341
+        '''
+        <div id="video-106908"
+            class="simplestream"
+            data-id="GB001"
+            data-type="vod"
+            data-key="3Li3Nt2Qs8Ct3Xq9Fi5Uy0Mb2Bj0Qs"
+            data-token="f9c317c727dc07f515b20036c8ef14a6"
+            data-expiry="1624300052"
+            data-uvid="37900558"
+            data-poster="https://thumbnails.simplestreamcdn.com/gbnews/ondemand/37900558.jpg?width=700&"
+            data-npaw="false"
+            data-env="production">
+        '''
+        # exception if no match
+        video_data = self._search_regex(
+            r'(<div\s[^>]*\bclass\s*=\s*(\'|")(?!.*sidebar\b)simplestream(?:\s[\s\w$-]*)?\2[^>]*>)',
+            webpage, 'video data')
+
+        video_data = extract_attributes(video_data)
+        ss_id = video_data.get('data-id')
+        if not ss_id:
+            raise ExtractorError('Simplestream ID not found')
+
+        json_data = self._download_json(
+            self._SSMP_URL, display_id,
+            note='Downloading Simplestream JSON metadata',
+            errnote='Unable to download Simplestream JSON metadata',
+            query={
+                'id': ss_id,
+                'env': video_data.get('data-env', 'production'),
+            }, fatal=False)
+
+        meta_url = traverse_obj(json_data, ('response', 'api_hostname'))
+        if not meta_url:
+            raise ExtractorError('No API host found')
+
+        uvid = video_data['data-uvid']
+        dtype = video_data.get('data-type')
+        stream_data = self._download_json(
+            '%s/api/%s/stream/%s' % (meta_url, 'show' if dtype == 'vod' else dtype, uvid),
+            uvid,
+            query={
+                'key': video_data.get('data-key'),
+                'platform': self._PLATFORM,
+            },
+            headers={
+                'Token': video_data.get('data-token'),
+                'Token-Expiry': video_data.get('data-expiry'),
+                'Uvid': uvid,
+            }, fatal=False)
+
+        stream_url = traverse_obj(stream_data, (
+            'response', 'stream', T(url_or_none)))
+        if not stream_url:
+            raise ExtractorError('No stream data/URL')
+
+        # now known to be a dict
+        stream_data = stream_data['response']
+        drm = stream_data.get('drm')
+        if drm:
+            self.report_drm(uvid)
+
+        formats = self._extract_m3u8_formats(
+            stream_url, uvid, ext='mp4', entry_protocol='m3u8_native',
+            fatal=False)
+        # exception if no formats
+        self._sort_formats(formats)
+
+        return {
+            'id': uvid,
+            'display_id': display_id,
+            'title': (traverse_obj(stream_data, ('title', T(txt_or_none)))
+                      or self._og_search_title(webpage, default=None)
+                      or display_id.replace('-', ' ').capitalize()),
+            'description': self._og_search_description(webpage, default=None),
+            'thumbnail': (traverse_obj(video_data, ('data-poster', T(url_or_none)))
+                          or self._og_search_thumbnail(webpage)),
+            'formats': formats,
+            'is_live': (dtype == 'live') or None,
+        }

youtube_dl/extractor/imgur.py

@@ -1,101 +1,267 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
 import re
 
 from .common import InfoExtractor
 from ..utils import (
+    determine_ext,
+    ExtractorError,
+    float_or_none,
     int_or_none,
     js_to_json,
+    merge_dicts,
     mimetype2ext,
-    ExtractorError,
+    parse_iso8601,
+    T,
+    traverse_obj,
+    txt_or_none,
+    url_or_none,
 )
 
 
-class ImgurIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:i\.)?imgur\.com/(?!(?:a|gallery|(?:t(?:opic)?|r)/[^/]+)/)(?P<id>[a-zA-Z0-9]+)'
+class ImgurBaseIE(InfoExtractor):
+    # hard-coded value, as also used by ArchiveTeam
+    _CLIENT_ID = '546c25a59c58ad7'
+
+    @classmethod
+    def _imgur_result(cls, item_id):
+        return cls.url_result('imgur:%s' % item_id, ImgurIE.ie_key(), item_id)
+
+    def _call_api(self, endpoint, video_id, **kwargs):
+        return self._download_json(
+            'https://api.imgur.com/post/v1/%s/%s?client_id=%s&include=media,account' % (endpoint, video_id, self._CLIENT_ID),
+            video_id, **kwargs)
+
+    @staticmethod
+    def get_description(s):
+        if 'Discover the magic of the internet at Imgur' in s:
+            return None
+        return txt_or_none(s)
+
+
+class ImgurIE(ImgurBaseIE):
+    _VALID_URL = r'''(?x)
+        (?:
+            https?://(?:i\.)?imgur\.com/(?!(?:a|gallery|t|topic|r)/)|
+            imgur:
+        )(?P<id>[a-zA-Z0-9]+)
+    '''
 
     _TESTS = [{
-        'url': 'https://i.imgur.com/A61SaA1.gifv',
+        'url': 'https://imgur.com/A61SaA1',
         'info_dict': {
             'id': 'A61SaA1',
             'ext': 'mp4',
             'title': 're:Imgur GIF$|MRW gifv is up and running without any bugs$',
+            'timestamp': 1416446068,
+            'upload_date': '20141120',
         },
     }, {
-        'url': 'https://imgur.com/A61SaA1',
+        'url': 'https://i.imgur.com/A61SaA1.gifv',
         'only_matching': True,
     }, {
         'url': 'https://i.imgur.com/crGpqCV.mp4',
         'only_matching': True,
     }, {
-        # no title
+        # previously, no title
         'url': 'https://i.imgur.com/jxBXAMC.gifv',
-        'only_matching': True,
+        'info_dict': {
+            'id': 'jxBXAMC',
+            'ext': 'mp4',
+            'title': 'Fahaka puffer feeding',
+            'timestamp': 1533835503,
+            'upload_date': '20180809',
+        },
     }]
 
+    def _extract_twitter_formats(self, html, tw_id='twitter', **kwargs):
+        fatal = kwargs.pop('fatal', False)
+        tw_stream = self._html_search_meta('twitter:player:stream', html, fatal=fatal, **kwargs)
+        if not tw_stream:
+            return []
+        ext = mimetype2ext(self._html_search_meta(
+            'twitter:player:stream:content_type', html, default=None))
+        width, height = (int_or_none(self._html_search_meta('twitter:player:' + v, html, default=None))
+                         for v in ('width', 'height'))
+        return [{
+            'format_id': tw_id,
+            'url': tw_stream,
+            'ext': ext or determine_ext(tw_stream),
+            'width': width,
+            'height': height,
+        }]
+
     def _real_extract(self, url):
         video_id = self._match_id(url)
+        data = self._call_api('media', video_id, fatal=False, expected_status=404)
         webpage = self._download_webpage(
-            'https://i.imgur.com/{id}.gifv'.format(id=video_id), video_id)
+            'https://i.imgur.com/{id}.gifv'.format(id=video_id), video_id, fatal=not data) or ''
 
-        width = int_or_none(self._og_search_property(
-            'video:width', webpage, default=None))
-        height = int_or_none(self._og_search_property(
-            'video:height', webpage, default=None))
+        if not traverse_obj(data, ('media', 0, (
+                ('type', T(lambda t: t == 'video' or None)),
+                ('metadata', 'is_animated'))), get_all=False):
+            raise ExtractorError(
+                '%s is not a video or animated image' % video_id,
+                expected=True)
+
+        media_fmt = traverse_obj(data, ('media', 0, {
+            'url': ('url', T(url_or_none)),
+            'ext': 'ext',
+            'width': ('width', T(int_or_none)),
+            'height': ('height', T(int_or_none)),
+            'filesize': ('size', T(int_or_none)),
+            'acodec': ('metadata', 'has_sound', T(lambda b: None if b else 'none')),
+        }))
+
+        media_url = traverse_obj(media_fmt, 'url')
+        if media_url:
+            if not media_fmt.get('ext'):
+                media_fmt['ext'] = mimetype2ext(traverse_obj(
+                    data, ('media', 0, 'mime_type'))) or determine_ext(media_url)
+            if traverse_obj(data, ('media', 0, 'type')) == 'image':
+                media_fmt['acodec'] = 'none'
+                media_fmt.setdefault('preference', -10)
+
+        tw_formats = self._extract_twitter_formats(webpage)
+        if traverse_obj(tw_formats, (0, 'url')) == media_url:
+            tw_formats = []
+        else:
+            # maybe this isn't an animated image/video?
+            self._check_formats(tw_formats, video_id)
 
         video_elements = self._search_regex(
             r'(?s)<div class="video-elements">(.*?)</div>',
             webpage, 'video elements', default=None)
-        if not video_elements:
+        if not (video_elements or tw_formats or media_url):
             raise ExtractorError(
-                'No sources found for video %s. Maybe an image?' % video_id,
+                'No sources found for video %s. Maybe a plain image?' % video_id,
                 expected=True)
 
-        formats = []
-        for m in re.finditer(r'<source\s+src="(?P<src>[^"]+)"\s+type="(?P<type>[^"]+)"', video_elements):
-            formats.append({
-                'format_id': m.group('type').partition('/')[2],
-                'url': self._proto_relative_url(m.group('src')),
-                'ext': mimetype2ext(m.group('type')),
-                'width': width,
-                'height': height,
+        def mung_format(fmt, *extra):
+            fmt.update({
                 'http_headers': {
                     'User-Agent': 'youtube-dl (like wget)',
                 },
             })
+            for d in extra:
+                fmt.update(d)
+            return fmt
 
-        gif_json = self._search_regex(
-            r'(?s)var\s+videoItem\s*=\s*(\{.*?\})',
-            webpage, 'GIF code', fatal=False)
-        if gif_json:
-            gifd = self._parse_json(
-                gif_json, video_id, transform_source=js_to_json)
-            formats.append({
-                'format_id': 'gif',
-                'preference': -10,
-                'width': width,
-                'height': height,
-                'ext': 'gif',
-                'acodec': 'none',
-                'vcodec': 'gif',
-                'container': 'gif',
-                'url': self._proto_relative_url(gifd['gifUrl']),
-                'filesize': gifd.get('size'),
-                'http_headers': {
-                    'User-Agent': 'youtube-dl (like wget)',
-                },
-            })
+        if video_elements:
+            def og_get_size(media_type):
+                return dict((p, int_or_none(self._og_search_property(
+                    ':'.join((media_type, p)), webpage, default=None)))
+                    for p in ('width', 'height'))
+
+            size = og_get_size('video')
+            if all(v is None for v in size.values()):
+                size = og_get_size('image')
+
+            formats = traverse_obj(
+                re.finditer(r'<source\s+src="(?P<src>[^"]+)"\s+type="(?P<type>[^"]+)"', video_elements),
+                (Ellipsis, {
+                    'format_id': ('type', T(lambda s: s.partition('/')[2])),
+                    'url': ('src', T(self._proto_relative_url)),
+                    'ext': ('type', T(mimetype2ext)),
+                }, T(lambda f: mung_format(f, size))))
+
+            gif_json = self._search_regex(
+                r'(?s)var\s+videoItem\s*=\s*(\{.*?\})',
+                webpage, 'GIF code', fatal=False)
+            MUST_BRANCH = (None, T(lambda _: None))
+            formats.extend(traverse_obj(gif_json, (
+                T(lambda j: self._parse_json(
+                    j, video_id, transform_source=js_to_json, fatal=False)), {
+                        'url': ('gifUrl', T(self._proto_relative_url)),
+                        'filesize': ('size', T(int_or_none)),
+                }, T(lambda f: mung_format(f, size, {
+                    'format_id': 'gif',
+                    'preference': -10,  # gifs are worse than videos
+                    'ext': 'gif',
+                    'acodec': 'none',
+                    'vcodec': 'gif',
+                    'container': 'gif',
+                })), MUST_BRANCH)))
+        else:
+            formats = []
+
+        # maybe add formats from JSON or page Twitter metadata
+        if not any((u == media_url) for u in traverse_obj(formats, (Ellipsis, 'url'))):
+            formats.append(mung_format(media_fmt))
+        tw_url = traverse_obj(tw_formats, (0, 'url'))
+        if not any((u == tw_url) for u in traverse_obj(formats, (Ellipsis, 'url'))):
+            formats.extend(mung_format(f) for f in tw_formats)
 
         self._sort_formats(formats)
 
-        return {
+        return merge_dicts(traverse_obj(data, {
+            'uploader_id': ('account_id', T(txt_or_none),
+                            T(lambda a: a if int_or_none(a) != 0 else None)),
+            'uploader': ('account', 'username', T(txt_or_none)),
+            'uploader_url': ('account', 'avatar_url', T(url_or_none)),
+            'like_count': ('upvote_count', T(int_or_none)),
+            'dislike_count': ('downvote_count', T(int_or_none)),
+            'comment_count': ('comment_count', T(int_or_none)),
+            'age_limit': ('is_mature', T(lambda x: 18 if x else None)),
+            'timestamp': (('updated_at', 'created_at'), T(parse_iso8601)),
+            'release_timestamp': ('created_at', T(parse_iso8601)),
+        }, get_all=False), traverse_obj(data, ('media', 0, 'metadata', {
+            'title': ('title', T(txt_or_none)),
+            'description': ('description', T(self.get_description)),
+            'duration': ('duration', T(float_or_none)),
+            'timestamp': (('updated_at', 'created_at'), T(parse_iso8601)),
+            'release_timestamp': ('created_at', T(parse_iso8601)),
+        })), {
             'id': video_id,
             'formats': formats,
-            'title': self._og_search_title(webpage, default=video_id),
-        }
+            'title': self._og_search_title(webpage, default='Imgur video ' + video_id),
+            'description': self.get_description(self._og_search_description(webpage)),
+            'thumbnail': url_or_none(self._html_search_meta('thumbnailUrl', webpage, default=None)),
+        })
 
 
-class ImgurGalleryIE(InfoExtractor):
+class ImgurGalleryBaseIE(ImgurBaseIE):
+    _GALLERY = True
+
+    def _real_extract(self, url):
+        gallery_id = self._match_id(url)
+
+        data = self._call_api('albums', gallery_id, fatal=False, expected_status=404)
+
+        info = traverse_obj(data, {
+            'title': ('title', T(txt_or_none)),
+            'description': ('description', T(self.get_description)),
+        })
+
+        if traverse_obj(data, 'is_album'):
+
+            def yield_media_ids():
+                for m_id in traverse_obj(data, (
+                        'media', lambda _, v: v.get('type') == 'video' or v['metadata']['is_animated'],
+                        'id', T(txt_or_none))):
+                    yield m_id
+
+            # if a gallery with exactly one video, apply album metadata to video
+            media_id = (
+                self._GALLERY
+                and traverse_obj(data, ('image_count', T(lambda c: c == 1)))
+                and next(yield_media_ids(), None))
+
+            if not media_id:
+                result = self.playlist_result(
+                    map(self._imgur_result, yield_media_ids()), gallery_id)
+                result.update(info)
+                return result
+            gallery_id = media_id
+
+        result = self._imgur_result(gallery_id)
+        info['_type'] = 'url_transparent'
+        result.update(info)
+        return result
+
+
+class ImgurGalleryIE(ImgurGalleryBaseIE):
     IE_NAME = 'imgur:gallery'
     _VALID_URL = r'https?://(?:i\.)?imgur\.com/(?:gallery|(?:t(?:opic)?|r)/[^/]+)/(?P<id>[a-zA-Z0-9]+)'
 
@@ -106,49 +272,93 @@ class ImgurGalleryIE(InfoExtractor):
             'title': 'Adding faces make every GIF better',
         },
         'playlist_count': 25,
+        'skip': 'Zoinks! You\'ve taken a wrong turn.',
     }, {
+        # TODO: static images - replace with animated/video gallery
         'url': 'http://imgur.com/topic/Aww/ll5Vk',
         'only_matching': True,
     }, {
         'url': 'https://imgur.com/gallery/YcAQlkx',
+        'add_ies': ['Imgur'],
         'info_dict': {
             'id': 'YcAQlkx',
             'ext': 'mp4',
             'title': 'Classic Steve Carell gif...cracks me up everytime....damn the repost downvotes....',
-        }
+            'timestamp': 1358554297,
+            'upload_date': '20130119',
+            'uploader_id': '1648642',
+            'uploader': 'wittyusernamehere',
+        },
     }, {
+        # TODO: static image - replace with animated/video gallery
         'url': 'http://imgur.com/topic/Funny/N8rOudd',
         'only_matching': True,
     }, {
         'url': 'http://imgur.com/r/aww/VQcQPhM',
-        'only_matching': True,
+        'add_ies': ['Imgur'],
+        'info_dict': {
+            'id': 'VQcQPhM',
+            'ext': 'mp4',
+            'title': 'The boss is here',
+            'timestamp': 1476494751,
+            'upload_date': '20161015',
+            'uploader_id': '19138530',
+            'uploader': 'thematrixcam',
+        },
+    },
+        # from PR #16674
+        {
+        'url': 'https://imgur.com/t/unmuted/6lAn9VQ',
+        'info_dict': {
+            'id': '6lAn9VQ',
+            'title': 'Penguins !',
+        },
+        'playlist_count': 3,
+    }, {
+        'url': 'https://imgur.com/t/unmuted/kx2uD3C',
+        'add_ies': ['Imgur'],
+        'info_dict': {
+            'id': 'ZVMv45i',
+            'ext': 'mp4',
+            'title': 'Intruder',
+            'timestamp': 1528129683,
+            'upload_date': '20180604',
+        },
+    }, {
+        'url': 'https://imgur.com/t/unmuted/wXSK0YH',
+        'add_ies': ['Imgur'],
+        'info_dict': {
+            'id': 'JCAP4io',
+            'ext': 'mp4',
+            'title': 're:I got the blues$',
+            'description': 'Luka’s vocal stylings.\n\nFP edit: don’t encourage me. I’ll never stop posting Luka and friends.',
+            'timestamp': 1527809525,
+            'upload_date': '20180531',
+        },
     }]
 
-    def _real_extract(self, url):
-        gallery_id = self._match_id(url)
 
-        data = self._download_json(
-            'https://imgur.com/gallery/%s.json' % gallery_id,
-            gallery_id)['data']['image']
-
-        if data.get('is_album'):
-            entries = [
-                self.url_result('http://imgur.com/%s' % image['hash'], ImgurIE.ie_key(), image['hash'])
-                for image in data['album_images']['images'] if image.get('hash')]
-            return self.playlist_result(entries, gallery_id, data.get('title'), data.get('description'))
-
-        return self.url_result('http://imgur.com/%s' % gallery_id, ImgurIE.ie_key(), gallery_id)
-
-
-class ImgurAlbumIE(ImgurGalleryIE):
+class ImgurAlbumIE(ImgurGalleryBaseIE):
     IE_NAME = 'imgur:album'
     _VALID_URL = r'https?://(?:i\.)?imgur\.com/a/(?P<id>[a-zA-Z0-9]+)'
-
+    _GALLERY = False
     _TESTS = [{
+        # TODO: only static images - replace with animated/video gallery
         'url': 'http://imgur.com/a/j6Orj',
+        'only_matching': True,
+    },
+        # from PR #21693
+        {
+        'url': 'https://imgur.com/a/iX265HX',
         'info_dict': {
-            'id': 'j6Orj',
-            'title': 'A Literary Analysis of "Star Wars: The Force Awakens"',
+            'id': 'iX265HX',
+            'title': 'enen-no-shouboutai'
         },
-        'playlist_count': 12,
+        'playlist_count': 2,
+    }, {
+        'url': 'https://imgur.com/a/8pih2Ed',
+        'info_dict': {
+            'id': '8pih2Ed'
+        },
+        'playlist_mincount': 1,
     }]

youtube_dl/extractor/telewebion.py

@@ -3,17 +3,23 @@ from __future__ import unicode_literals
 
 from .common import InfoExtractor
 
+from ..utils import (
+    float_or_none,
+    int_or_none,
+    url_or_none,
+)
+
 
 class TelewebionIE(InfoExtractor):
-    _VALID_URL = r'https?://(?:www\.)?telewebion\.com/#!/episode/(?P<id>\d+)'
+    _VALID_URL = r'https?://(?:www\.)?telewebion\.com/(episode|clip)/(?P<id>[a-zA-Z0-9]+)'
 
     _TEST = {
-        'url': 'http://www.telewebion.com/#!/episode/1263668/',
+        'url': 'http://www.telewebion.com/episode/0x1b3139c/',
         'info_dict': {
-            'id': '1263668',
+            'id': '0x1b3139c',
             'ext': 'mp4',
             'title': 'قرعه\u200cکشی لیگ قهرمانان اروپا',
-            'thumbnail': r're:^https?://.*\.jpg',
+            'thumbnail': r're:^https?://static\.telewebion\.com/episodeImages/.*/default',
             'view_count': int,
         },
         'params': {
@@ -25,31 +31,24 @@ class TelewebionIE(InfoExtractor):
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        secure_token = self._download_webpage(
-            'http://m.s2.telewebion.com/op/op?action=getSecurityToken', video_id)
-        episode_details = self._download_json(
-            'http://m.s2.telewebion.com/op/op', video_id,
-            query={'action': 'getEpisodeDetails', 'episode_id': video_id})
+        episode_details = self._download_json('https://gateway.telewebion.ir/kandoo/episode/getEpisodeDetail/?EpisodeId={0}'.format(video_id), video_id)
+        episode_details = episode_details['body']['queryEpisode'][0]
 
-        m3u8_url = 'http://m.s1.telewebion.com/smil/%s.m3u8?filepath=%s&m3u8=1&secure_token=%s' % (
-            video_id, episode_details['file_path'], secure_token)
+        channel_id = episode_details['channel']['descriptor']
+        episode_image_id = episode_details.get('image')
+        episode_image = 'https://static.telewebion.com/episodeImages/{0}/default'.format(episode_image_id) if episode_image_id else None
+
+        m3u8_url = 'https://cdna.telewebion.com/{0}/episode/{1}/playlist.m3u8'.format(channel_id, video_id)
         formats = self._extract_m3u8_formats(
-            m3u8_url, video_id, ext='mp4', m3u8_id='hls')
-
-        picture_paths = [
-            episode_details.get('picture_path'),
-            episode_details.get('large_picture_path'),
-        ]
-
-        thumbnails = [{
-            'url': picture_path,
-            'preference': idx,
-        } for idx, picture_path in enumerate(picture_paths) if picture_path is not None]
+            m3u8_url, video_id, ext='mp4', m3u8_id='hls',
+            entry_protocol='m3u8_native')
+        self._sort_formats(formats)
 
         return {
             'id': video_id,
             'title': episode_details['title'],
             'formats': formats,
-            'thumbnails': thumbnails,
-            'view_count': episode_details.get('view_count'),
+            'thumbnail': url_or_none(episode_image),
+            'view_count': int_or_none(episode_details.get('view_count')),
+            'duration': float_or_none(episode_details.get('duration')),
         }

youtube_dl/extractor/vbox7.py

@@ -2,9 +2,22 @@
 from __future__ import unicode_literals
 
 import re
+import time
 
 from .common import InfoExtractor
-from ..utils import ExtractorError
+from ..compat import compat_kwargs
+from ..utils import (
+    base_url,
+    determine_ext,
+    ExtractorError,
+    float_or_none,
+    merge_dicts,
+    T,
+    traverse_obj,
+    txt_or_none,
+    url_basename,
+    url_or_none,
+)
 
 
 class Vbox7IE(InfoExtractor):
@@ -20,23 +33,27 @@ class Vbox7IE(InfoExtractor):
                         )
                         (?P<id>[\da-fA-F]+)
                     '''
+    _EMBED_REGEX = [r'<iframe[^>]+src=(?P<q>["\'])(?P<url>(?:https?:)?//vbox7\.com/emb/external\.php.+?)(?P=q)']
     _GEO_COUNTRIES = ['BG']
     _TESTS = [{
-        'url': 'http://vbox7.com/play:0946fff23c',
-        'md5': 'a60f9ab3a3a2f013ef9a967d5f7be5bf',
+        # the http: URL just redirects here
+        'url': 'https://vbox7.com/play:0946fff23c',
+        'md5': '50ca1f78345a9c15391af47d8062d074',
         'info_dict': {
             'id': '0946fff23c',
             'ext': 'mp4',
             'title': 'Борисов: Притеснен съм за бъдещето на България',
             'description': 'По думите му е опасно страната ни да бъде обявена за "сигурна"',
-            'thumbnail': r're:^https?://.*\.jpg$',
             'timestamp': 1470982814,
             'upload_date': '20160812',
             'uploader': 'zdraveibulgaria',
+            'thumbnail': r're:^https?://.*\.jpg$',
+            'view_count': int,
+            'duration': 2640,
         },
-        'params': {
-            'proxy': '127.0.0.1:8118',
-        },
+        'expected_warnings': [
+            'Unable to download webpage',
+        ],
     }, {
         'url': 'http://vbox7.com/play:249bb972c2',
         'md5': '99f65c0c9ef9b682b97313e052734c3f',
@@ -44,8 +61,15 @@ class Vbox7IE(InfoExtractor):
             'id': '249bb972c2',
             'ext': 'mp4',
             'title': 'Смях! Чудо - чист за секунди - Скрита камера',
+            'description': 'Смях! Чудо - чист за секунди - Скрита камера',
+            'timestamp': 1360215023,
+            'upload_date': '20130207',
+            'uploader': 'svideteliat_ot_varshava',
+            'thumbnail': 'https://i49.vbox7.com/o/249/249bb972c20.jpg',
+            'view_count': int,
+            'duration': 83,
         },
-        'skip': 'georestricted',
+        'expected_warnings': ['Failed to download m3u8 information'],
     }, {
         'url': 'http://vbox7.com/emb/external.php?vid=a240d20f9c&autoplay=1',
         'only_matching': True,
@@ -54,52 +78,127 @@ class Vbox7IE(InfoExtractor):
         'only_matching': True,
     }]
 
-    @staticmethod
-    def _extract_url(webpage):
-        mobj = re.search(
-            r'<iframe[^>]+src=(?P<q>["\'])(?P<url>(?:https?:)?//vbox7\.com/emb/external\.php.+?)(?P=q)',
-            webpage)
+    @classmethod
+    def _extract_url(cls, webpage):
+        mobj = re.search(cls._EMBED_REGEX[0], webpage)
         if mobj:
             return mobj.group('url')
 
+    # specialisation to transform what looks like ld+json that
+    # may contain invalid character combinations
+
+    # transform_source=None, fatal=True
+    def _parse_json(self, json_string, video_id, *args, **kwargs):
+        if '"@context"' in json_string[:30]:
+            # this is ld+json, or that's the way to bet
+            transform_source = args[0] if len(args) > 0 else kwargs.get('transform_source')
+            if not transform_source:
+
+                def fix_chars(src):
+                    # fix malformed ld+json: replace raw CRLFs with escaped LFs
+                    return re.sub(
+                        r'"[^"]+"', lambda m: re.sub(r'\r?\n', r'\\n', m.group(0)), src)
+
+                if len(args) > 0:
+                    args = (fix_chars,) + args[1:]
+                else:
+                    kwargs['transform_source'] = fix_chars
+                    kwargs = compat_kwargs(kwargs)
+
+        return super(Vbox7IE, self)._parse_json(
+            json_string, video_id, *args, **kwargs)
+
     def _real_extract(self, url):
         video_id = self._match_id(url)
+        url = 'https://vbox7.com/play:%s' % (video_id,)
 
+        now = time.time()
         response = self._download_json(
-            'https://www.vbox7.com/ajax/video/nextvideo.php?vid=%s' % video_id,
-            video_id)
+            'https://www.vbox7.com/aj/player/item/options', video_id,
+            query={'vid': video_id}, headers={'Referer': url})
+        # estimate time to which possible `ago` member is relative
+        now = now + 0.5 * (time.time() - now)
 
-        if 'error' in response:
+        if traverse_obj(response, 'error'):
             raise ExtractorError(
                 '%s said: %s' % (self.IE_NAME, response['error']), expected=True)
 
-        video = response['options']
+        src_url = traverse_obj(response, ('options', 'src', T(url_or_none))) or ''
 
-        title = video['title']
-        video_url = video['src']
-
-        if '/na.mp4' in video_url:
+        fmt_base = url_basename(src_url).rsplit('.', 1)[0].rsplit('_', 1)[0]
+        if fmt_base in ('na', 'vn'):
             self.raise_geo_restricted(countries=self._GEO_COUNTRIES)
 
-        uploader = video.get('uploader')
+        ext = determine_ext(src_url)
+        if ext == 'mpd':
+            # extract MPD
+            try:
+                formats, subtitles = self._extract_mpd_formats_and_subtitles(
+                    src_url, video_id, 'dash', fatal=False)
+            except KeyError:  # fatal doesn't catch this
+                self.report_warning('Failed to parse MPD manifest')
+                formats, subtitles = [], {}
+        elif ext != 'm3u8':
+            formats = [{
+                'url': src_url,
+            }] if src_url else []
+            subtitles = {}
 
-        webpage = self._download_webpage(
-            'http://vbox7.com/play:%s' % video_id, video_id, fatal=None)
+        if src_url:
+            # possibly extract HLS, based on https://github.com/yt-dlp/yt-dlp/pull/9100
+            fmt_base = base_url(src_url) + fmt_base
+            # prepare for _extract_m3u8_formats_and_subtitles()
+            # hls_formats, hls_subs = self._extract_m3u8_formats_and_subtitles(
+            hls_formats = self._extract_m3u8_formats(
+                '{0}.m3u8'.format(fmt_base), video_id, m3u8_id='hls', fatal=False)
+            formats.extend(hls_formats)
+            # self._merge_subtitles(hls_subs, target=subtitles)
 
-        info = {}
+            # In case MPD/HLS cannot be parsed, or anyway, get mp4 combined
+            # formats usually provided to Safari, iOS, and old Windows
+            video = response['options']
+            resolutions = (1080, 720, 480, 240, 144)
+            highest_res = traverse_obj(video, (
+                'highestRes', T(int))) or resolutions[0]
+            resolutions = traverse_obj(video, (
+                'resolutions', lambda _, r: highest_res >= int(r) > 0)) or resolutions
+            mp4_formats = traverse_obj(resolutions, (
+                Ellipsis, T(lambda res: {
+                    'url': '{0}_{1}.mp4'.format(fmt_base, res),
+                    'format_id': 'http-{0}'.format(res),
+                    'height': res,
+                })))
+            # if above formats are flaky, enable the line below
+            # self._check_formats(mp4_formats, video_id)
+            formats.extend(mp4_formats)
 
-        if webpage:
-            info = self._search_json_ld(
-                webpage.replace('"/*@context"', '"@context"'), video_id,
-                fatal=False)
+        self._sort_formats(formats)
 
-        info.update({
+        webpage = self._download_webpage(url, video_id, fatal=False) or ''
+
+        info = self._search_json_ld(
+            webpage.replace('"/*@context"', '"@context"'), video_id,
+            fatal=False) if webpage else {}
+
+        if not info.get('title'):
+            info['title'] = traverse_obj(response, (
+                'options', 'title', T(txt_or_none))) or self._og_search_title(webpage)
+
+        def if_missing(k):
+            return lambda x: None if k in info else x
+
+        info = merge_dicts(info, {
             'id': video_id,
-            'title': title,
-            'url': video_url,
-            'uploader': uploader,
-            'thumbnail': self._proto_relative_url(
+            'formats': formats,
+            'subtitles': subtitles or None,
+        }, info, traverse_obj(response, ('options', {
+            'uploader': ('uploader', T(txt_or_none)),
+            'timestamp': ('ago', T(if_missing('timestamp')), T(lambda t: int(round((now - t) / 60.0)) * 60)),
+            'duration': ('duration', T(if_missing('duration')), T(float_or_none)),
+        })))
+        if 'thumbnail' not in info:
+            info['thumbnail'] = self._proto_relative_url(
                 info.get('thumbnail') or self._og_search_thumbnail(webpage),
-                'http:'),
-        })
+                'https:'),
+
         return info

youtube_dl/extractor/videa.py

@@ -6,22 +6,31 @@ import re
 import string
 
 from .common import InfoExtractor
+from ..compat import (
+    compat_b64decode,
+    compat_ord,
+    compat_struct_pack,
+)
 from ..utils import (
     ExtractorError,
     int_or_none,
     mimetype2ext,
     parse_codecs,
+    parse_qs,
     update_url_query,
     urljoin,
     xpath_element,
     xpath_text,
 )
-from ..compat import (
-    compat_b64decode,
-    compat_ord,
-    compat_struct_pack,
-    compat_urlparse,
-)
+
+
+def compat_random_choices(population, *args, **kwargs):
+    # weights=None, *, cum_weights=None, k=1
+    # limited implementation needed here
+    weights = args[0] if args else kwargs.get('weights')
+    assert all(w is None for w in (weights, kwargs.get('cum_weights')))
+    k = kwargs.get('k', 1)
+    return ''.join(random.choice(population) for _ in range(k))
 
 
 class VideaIE(InfoExtractor):
@@ -35,6 +44,7 @@ class VideaIE(InfoExtractor):
                         )
                         (?P<id>[^?#&]+)
                     '''
+    _EMBED_REGEX = [r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//videa\.hu/player\?.*?\bv=.+?)\1']
     _TESTS = [{
         'url': 'http://videa.hu/videok/allatok/az-orult-kigyasz-285-kigyot-kigyo-8YfIAjxwWGwT8HVQ',
         'md5': '97a7af41faeaffd9f1fc864a7c7e7603',
@@ -44,6 +54,7 @@ class VideaIE(InfoExtractor):
             'title': 'Az őrült kígyász 285 kígyót enged szabadon',
             'thumbnail': r're:^https?://.*',
             'duration': 21,
+            'age_limit': 0,
         },
     }, {
         'url': 'http://videa.hu/videok/origo/jarmuvek/supercars-elozes-jAHDWfWSJH5XuFhH',
@@ -54,6 +65,7 @@ class VideaIE(InfoExtractor):
             'title': 'Supercars előzés',
             'thumbnail': r're:^https?://.*',
             'duration': 64,
+            'age_limit': 0,
         },
     }, {
         'url': 'http://videa.hu/player?v=8YfIAjxwWGwT8HVQ',
@@ -64,6 +76,7 @@ class VideaIE(InfoExtractor):
             'title': 'Az őrült kígyász 285 kígyót enged szabadon',
             'thumbnail': r're:^https?://.*',
             'duration': 21,
+            'age_limit': 0,
         },
     }, {
         'url': 'http://videa.hu/player/v/8YfIAjxwWGwT8HVQ?autoplay=1',
@@ -80,11 +93,14 @@ class VideaIE(InfoExtractor):
     }]
     _STATIC_SECRET = 'xHb0ZvME5q8CBcoQi6AngerDu3FGO9fkUlwPmLVY_RTzj2hJIS4NasXWKy1td7p'
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return [url for _, url in re.findall(
-            r'<iframe[^>]+src=(["\'])(?P<url>(?:https?:)?//videa\.hu/player\?.*?\bv=.+?)\1',
-            webpage)]
+    @classmethod
+    def _extract_urls(cls, webpage):
+        def yield_urls():
+            for pattern in cls._EMBED_REGEX:
+                for m in re.finditer(pattern, webpage):
+                    yield m.group('url')
+
+        return list(yield_urls())
 
     @staticmethod
     def rc4(cipher_text, key):
@@ -130,13 +146,13 @@ class VideaIE(InfoExtractor):
         for i in range(0, 32):
             result += s[i - (self._STATIC_SECRET.index(l[i]) - 31)]
 
-        query = compat_urlparse.parse_qs(compat_urlparse.urlparse(player_url).query)
-        random_seed = ''.join(random.choice(string.ascii_letters + string.digits) for _ in range(8))
+        query = parse_qs(player_url)
+        random_seed = ''.join(compat_random_choices(string.ascii_letters + string.digits, k=8))
         query['_s'] = random_seed
         query['_t'] = result[:16]
 
         b64_info, handle = self._download_webpage_handle(
-            'http://videa.hu/videaplayer_get_xml.php', video_id, query=query)
+            'http://videa.hu/player/xml', video_id, query=query)
         if b64_info.startswith('<?xml'):
             info = self._parse_xml(b64_info, video_id)
         else:

youtube_dl/extractor/vimeo.py

@@ -673,8 +673,8 @@ class VimeoIE(VimeoBaseInfoExtractor):
             raise
 
         if '//player.vimeo.com/video/' in url:
-            config = self._parse_json(self._search_regex(
-                r'(?s)\b(?:playerC|c)onfig\s*=\s*({.+?})\s*[;\n]', webpage, 'info section'), video_id)
+            config = self._search_json(
+                r'\b(?:playerC|c)onfig\s*=', webpage, 'info section', video_id)
             if config.get('view') == 4:
                 config = self._verify_player_video_password(
                     redirect_url, video_id, headers)

youtube_dl/extractor/xfileshare.py

@@ -4,20 +4,28 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
-from ..compat import compat_chr
+from ..compat import (
+    compat_chr,
+    compat_zip as zip,
+)
 from ..utils import (
+    clean_html,
     decode_packed_codes,
     determine_ext,
     ExtractorError,
+    get_element_by_id,
     int_or_none,
-    js_to_json,
+    merge_dicts,
+    T,
+    traverse_obj,
+    url_or_none,
     urlencode_postdata,
 )
 
 
 # based on openload_decode from 2bfeee69b976fe049761dd3012e30b637ee05a58
 def aa_decode(aa_code):
-    symbol_table = [
+    symbol_table = (
         ('7', '((ﾟｰﾟ) + (o^_^o))'),
         ('6', '((o^_^o) +(o^_^o))'),
         ('5', '((ﾟｰﾟ) + (ﾟΘﾟ))'),
@@ -26,84 +34,180 @@ def aa_decode(aa_code):
         ('3', '(o^_^o)'),
         ('1', '(ﾟΘﾟ)'),
         ('0', '(c^_^o)'),
-    ]
+        ('+', ''),
+    )
     delim = '(ﾟДﾟ)[ﾟεﾟ]+'
-    ret = ''
-    for aa_char in aa_code.split(delim):
+
+    def chr_from_code(c):
         for val, pat in symbol_table:
-            aa_char = aa_char.replace(pat, val)
-        aa_char = aa_char.replace('+ ', '')
-        m = re.match(r'^\d+', aa_char)
-        if m:
-            ret += compat_chr(int(m.group(0), 8))
+            c = c.replace(pat, val)
+        if c.startswith(('u', 'U')):
+            base = 16
+            c = c[1:]
         else:
-            m = re.match(r'^u([\da-f]+)', aa_char)
-            if m:
-                ret += compat_chr(int(m.group(1), 16))
-    return ret
+            base = 10
+        c = int_or_none(c, base=base)
+        return '' if c is None else compat_chr(c)
+
+    return ''.join(
+        chr_from_code(aa_char)
+        for aa_char in aa_code.split(delim))
 
 
 class XFileShareIE(InfoExtractor):
     _SITES = (
-        (r'aparat\.cam', 'Aparat'),
-        (r'clipwatching\.com', 'ClipWatching'),
-        (r'gounlimited\.to', 'GoUnlimited'),
-        (r'govid\.me', 'GoVid'),
-        (r'holavid\.com', 'HolaVid'),
-        (r'streamty\.com', 'Streamty'),
-        (r'thevideobee\.to', 'TheVideoBee'),
-        (r'uqload\.com', 'Uqload'),
-        (r'vidbom\.com', 'VidBom'),
-        (r'vidlo\.us', 'vidlo'),
-        (r'vidlocker\.xyz', 'VidLocker'),
-        (r'vidshare\.tv', 'VidShare'),
-        (r'vup\.to', 'VUp'),
+        # status check 2024-02: site availability, G site: search
+        (r'aparat\.cam', 'Aparat'),  # Cloudflare says host error 522, apparently changed to wolfstreeam.tv
+        (r'filemoon\.sx/.', 'FileMoon'),
+        (r'gounlimited\.to', 'GoUnlimited'),  # no media pages listed
+        (r'govid\.me', 'GoVid'),  # no media pages listed
+        (r'highstream\.tv', 'HighStream'),  # clipwatching.com redirects here
+        (r'holavid\.com', 'HolaVid'),  # Cloudflare says host error 522
+        # (r'streamty\.com', 'Streamty'),  # no media pages listed, connection timeout
+        # (r'thevideobee\.to', 'TheVideoBee'),  # no pages listed, refuses connection
+        (r'uqload\.to', 'Uqload'),  # .com, .co redirect here
+        (r'(?:vedbam\.xyz|vadbam.net)', 'V?dB?m'),  # vidbom.com redirects here, but no valid media pages listed
+        (r'vidlo\.us', 'vidlo'),  # no valid media pages listed
+        (r'vidlocker\.xyz', 'VidLocker'),  # no media pages listed
+        (r'(?:w\d\.)?viidshar\.com', 'VidShare'),  # vidshare.tv redirects here
+        # (r'vup\.to', 'VUp'),  # domain not found
         (r'wolfstream\.tv', 'WolfStream'),
-        (r'xvideosharing\.com', 'XVideoSharing'),
+        (r'xvideosharing\.com', 'XVideoSharing'),  # just started showing 'maintenance mode'
     )
 
-    IE_DESC = 'XFileShare based sites: %s' % ', '.join(list(zip(*_SITES))[1])
+    IE_DESC = 'XFileShare-based sites: %s' % ', '.join(list(zip(*_SITES))[1])
     _VALID_URL = (r'https?://(?:www\.)?(?P<host>%s)/(?:embed-)?(?P<id>[0-9a-zA-Z]+)'
                   % '|'.join(site for site in list(zip(*_SITES))[0]))
+    _EMBED_REGEX = [r'<iframe\b[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//(?:%s)/embed-[0-9a-zA-Z]+.*?)\1' % '|'.join(site for site in list(zip(*_SITES))[0])]
 
     _FILE_NOT_FOUND_REGEXES = (
         r'>(?:404 - )?File Not Found<',
         r'>The file was removed by administrator<',
     )
+    _TITLE_REGEXES = (
+        r'style="z-index: [0-9]+;">([^<]+)</span>',
+        r'<td nowrap>([^<]+)</td>',
+        r'h4-fine[^>]*>([^<]+)<',
+        r'>Watch (.+)[ <]',
+        r'<h2 class="video-page-head">([^<]+)</h2>',
+        r'<h2 style="[^"]*color:#403f3d[^"]*"[^>]*>([^<]+)<',  # streamin.to (dead)
+        r'title\s*:\s*"([^"]+)"',  # govid.me
+    )
+    _SOURCE_URL_REGEXES = (
+        r'(?:file|src)\s*:\s*(["\'])(?P<url>http(?:(?!\1).)+\.(?:m3u8|mp4|flv)(?:(?!\1).)*)\1',
+        r'file_link\s*=\s*(["\'])(?P<url>http(?:(?!\1).)+)\1',
+        r'addVariable\((\\?["\'])file\1\s*,\s*(\\?["\'])(?P<url>http(?:(?!\2).)+)\2\)',
+        r'<embed[^>]+src=(["\'])(?P<url>http(?:(?!\1).)+\.(?:m3u8|mp4|flv)(?:(?!\1).)*)\1',
+    )
+    _THUMBNAIL_REGEXES = (
+        r'<video[^>]+poster="([^"]+)"',
+        r'(?:image|poster)\s*:\s*["\'](http[^"\']+)["\'],',
+    )
 
     _TESTS = [{
-        'url': 'http://xvideosharing.com/fq65f94nd2ve',
-        'md5': '4181f63957e8fe90ac836fa58dc3c8a6',
+        'note': 'link in `sources`',
+        'url': 'https://uqload.to/dcsu06gdb45o',
+        'md5': '7f8db187b254379440bf4fcad094ae86',
         'info_dict': {
-            'id': 'fq65f94nd2ve',
+            'id': 'dcsu06gdb45o',
             'ext': 'mp4',
-            'title': 'sample',
-            'thumbnail': r're:http://.*\.jpg',
+            'title': 'f2e31015957e74c8c8427982e161c3fc mp4',
+            'thumbnail': r're:https://.*\.jpg'
+        },
+        'params': {
+            'nocheckcertificate': True,
+        },
+        'expected_warnings': ['Unable to extract JWPlayer data'],
+    }, {
+        'note': 'link in decoded `sources`',
+        'url': 'https://xvideosharing.com/1tlg6agrrdgc',
+        'md5': '2608ce41932c1657ae56258a64e647d9',
+        'info_dict': {
+            'id': '1tlg6agrrdgc',
+            'ext': 'mp4',
+            'title': '0121',
+            'thumbnail': r're:https?://.*\.jpg',
+        },
+        'skip': 'This server is in maintenance mode.',
+    }, {
+        'note': 'JWPlayer link in un-p,a,c,k,e,d JS',
+        'url': 'https://filemoon.sx/e/dw40rxrzruqz',
+        'md5': '5a713742f57ac4aef29b74733e8dda01',
+        'info_dict': {
+            'id': 'dw40rxrzruqz',
+            'title': 'dw40rxrzruqz',
+            'ext': 'mp4'
+        },
+    }, {
+        'note': 'JWPlayer link in un-p,a,c,k,e,d JS',
+        'url': 'https://vadbam.net/6lnbkci96wly.html',
+        'md5': 'a1616800076177e2ac769203957c54bc',
+        'info_dict': {
+            'id': '6lnbkci96wly',
+            'title': 'Heart Crime S01 E03 weciima autos',
+            'ext': 'mp4'
+        },
+    }, {
+        'note': 'JWPlayer link in clear',
+        'url': 'https://w1.viidshar.com/nnibe0xf0h79.html',
+        'md5': 'f0a580ce9df06cc61b4a5c979d672367',
+        'info_dict': {
+            'id': 'nnibe0xf0h79',
+            'title': 'JaGa 68ar',
+            'ext': 'mp4'
+        },
+        'params': {
+            'skip_download': 'ffmpeg',
+        },
+        'expected_warnings': ['hlsnative has detected features it does not support'],
+    }, {
+        'note': 'JWPlayer link in clear',
+        'url': 'https://wolfstream.tv/a3drtehyrg52.html',
+        'md5': '1901d86a79c5e0c6a51bdc9a4cfd3769',
+        'info_dict': {
+            'id': 'a3drtehyrg52',
+            'title': 'NFL 2023 W04 DET@GB',
+            'ext': 'mp4'
         },
     }, {
         'url': 'https://aparat.cam/n4d6dh0wvlpr',
         'only_matching': True,
     }, {
-        'url': 'https://wolfstream.tv/nthme29v9u2x',
+        'url': 'https://uqload.to/ug5somm0ctnk.html',
+        'only_matching': True,
+    }, {
+        'url': 'https://highstream.tv/2owiyz3sjoux',
+        'only_matching': True,
+    }, {
+        'url': 'https://vedbam.xyz/6lnbkci96wly.html',
         'only_matching': True,
     }]
 
-    @staticmethod
-    def _extract_urls(webpage):
-        return [
-            mobj.group('url')
-            for mobj in re.finditer(
-                r'<iframe\b[^>]+\bsrc=(["\'])(?P<url>(?:https?:)?//(?:%s)/embed-[0-9a-zA-Z]+.*?)\1'
-                % '|'.join(site for site in list(zip(*XFileShareIE._SITES))[0]),
-                webpage)]
+    @classmethod
+    def _extract_urls(cls, webpage):
+
+        def yield_urls():
+            for regex in cls._EMBED_REGEX:
+                for mobj in re.finditer(regex, webpage):
+                    yield mobj.group('url')
+
+        return list(yield_urls())
 
     def _real_extract(self, url):
-        host, video_id = re.match(self._VALID_URL, url).groups()
+        host, video_id = self._match_valid_url(url).group('host', 'id')
 
-        url = 'https://%s/' % host + ('embed-%s.html' % video_id if host in ('govid.me', 'vidlo.us') else video_id)
+        url = 'https://%s/%s' % (
+            host,
+            'embed-%s.html' % video_id if host in ('govid.me', 'vidlo.us') else video_id)
         webpage = self._download_webpage(url, video_id)
-
-        if any(re.search(p, webpage) for p in self._FILE_NOT_FOUND_REGEXES):
+        container_div = get_element_by_id('container', webpage) or webpage
+        if self._search_regex(
+                r'>This server is in maintenance mode\.', container_div,
+                'maint error', group=0, default=None):
+            raise ExtractorError(clean_html(container_div), expected=True)
+        if self._search_regex(
+                self._FILE_NOT_FOUND_REGEXES, container_div,
+                'missing video error', group=0, default=None):
             raise ExtractorError('Video %s does not exist' % video_id, expected=True)
 
         fields = self._hidden_inputs(webpage)
@@ -122,59 +226,43 @@ class XFileShareIE(InfoExtractor):
                     'Content-type': 'application/x-www-form-urlencoded',
                 })
 
-        title = (self._search_regex(
-            (r'style="z-index: [0-9]+;">([^<]+)</span>',
-             r'<td nowrap>([^<]+)</td>',
-             r'h4-fine[^>]*>([^<]+)<',
-             r'>Watch (.+)[ <]',
-             r'<h2 class="video-page-head">([^<]+)</h2>',
-             r'<h2 style="[^"]*color:#403f3d[^"]*"[^>]*>([^<]+)<',  # streamin.to
-             r'title\s*:\s*"([^"]+)"'),  # govid.me
-            webpage, 'title', default=None) or self._og_search_title(
-            webpage, default=None) or video_id).strip()
+        title = (
+            self._search_regex(self._TITLE_REGEXES, webpage, 'title', default=None)
+            or self._og_search_title(webpage, default=None)
+            or video_id).strip()
 
-        for regex, func in (
-                (r'(eval\(function\(p,a,c,k,e,d\){.+)', decode_packed_codes),
-                (r'(ﾟ.+)', aa_decode)):
-            obf_code = self._search_regex(regex, webpage, 'obfuscated code', default=None)
-            if obf_code:
-                webpage = webpage.replace(obf_code, func(obf_code))
+        obf_code = True
+        while obf_code:
+            for regex, func in (
+                    (r'(?s)(?<!-)\b(eval\(function\(p,a,c,k,e,d\)\{(?:(?!</script>).)+\)\))',
+                     decode_packed_codes),
+                    (r'(ﾟ.+)', aa_decode)):
+                obf_code = self._search_regex(regex, webpage, 'obfuscated code', default=None)
+                if obf_code:
+                    webpage = webpage.replace(obf_code, func(obf_code))
+                    break
 
-        formats = []
+        jwplayer_data = self._find_jwplayer_data(
+            webpage.replace(r'\'', '\''), video_id)
+        result = self._parse_jwplayer_data(
+            jwplayer_data, video_id, require_title=False,
+            m3u8_id='hls', mpd_id='dash')
 
-        jwplayer_data = self._search_regex(
-            [
-                r'jwplayer\("[^"]+"\)\.load\(\[({.+?})\]\);',
-                r'jwplayer\("[^"]+"\)\.setup\(({.+?})\);',
-            ], webpage,
-            'jwplayer data', default=None)
-        if jwplayer_data:
-            jwplayer_data = self._parse_json(
-                jwplayer_data.replace(r"\'", "'"), video_id, js_to_json)
+        if not traverse_obj(result, 'formats'):
             if jwplayer_data:
-                formats = self._parse_jwplayer_data(
-                    jwplayer_data, video_id, False,
-                    m3u8_id='hls', mpd_id='dash')['formats']
-
-        if not formats:
-            urls = []
-            for regex in (
-                    r'(?:file|src)\s*:\s*(["\'])(?P<url>http(?:(?!\1).)+\.(?:m3u8|mp4|flv)(?:(?!\1).)*)\1',
-                    r'file_link\s*=\s*(["\'])(?P<url>http(?:(?!\1).)+)\1',
-                    r'addVariable\((\\?["\'])file\1\s*,\s*(\\?["\'])(?P<url>http(?:(?!\2).)+)\2\)',
-                    r'<embed[^>]+src=(["\'])(?P<url>http(?:(?!\1).)+\.(?:m3u8|mp4|flv)(?:(?!\1).)*)\1'):
+                self.report_warning(
+                    'Failed to extract JWPlayer formats', video_id=video_id)
+            urls = set()
+            for regex in self._SOURCE_URL_REGEXES:
                 for mobj in re.finditer(regex, webpage):
-                    video_url = mobj.group('url')
-                    if video_url not in urls:
-                        urls.append(video_url)
+                    urls.add(mobj.group('url'))
 
             sources = self._search_regex(
                 r'sources\s*:\s*(\[(?!{)[^\]]+\])', webpage, 'sources', default=None)
-            if sources:
-                urls.extend(self._parse_json(sources, video_id))
+            urls.update(traverse_obj(sources, (T(lambda s: self._parse_json(s, video_id)), Ellipsis)))
 
             formats = []
-            for video_url in urls:
+            for video_url in traverse_obj(urls, (Ellipsis, T(url_or_none))):
                 if determine_ext(video_url) == 'm3u8':
                     formats.extend(self._extract_m3u8_formats(
                         video_url, video_id, 'mp4',
@@ -185,17 +273,19 @@ class XFileShareIE(InfoExtractor):
                         'url': video_url,
                         'format_id': 'sd',
                     })
-        self._sort_formats(formats)
+            result = {'formats': formats}
+
+        self._sort_formats(result['formats'])
 
         thumbnail = self._search_regex(
-            [
-                r'<video[^>]+poster="([^"]+)"',
-                r'(?:image|poster)\s*:\s*["\'](http[^"\']+)["\'],',
-            ], webpage, 'thumbnail', default=None)
+            self._THUMBNAIL_REGEXES, webpage, 'thumbnail', default=None)
 
-        return {
+        if not (title or result.get('title')):
+            title = self._generic_title(url) or video_id
+
+        return merge_dicts(result, {
             'id': video_id,
-            'title': title,
+            'title': title or None,
             'thumbnail': thumbnail,
-            'formats': formats,
-        }
+            'http_headers': {'Referer': url}
+        })

youtube_dl/extractor/youtube.py

@@ -2,6 +2,7 @@
 
 from __future__ import unicode_literals
 
+import collections
 import itertools
 import json
 import os.path
@@ -23,10 +24,10 @@ from ..compat import (
 )
 from ..jsinterp import JSInterpreter
 from ..utils import (
-    ExtractorError,
     clean_html,
     dict_get,
     error_to_compat_str,
+    ExtractorError,
     float_or_none,
     extract_attributes,
     get_element_by_attribute,
@@ -36,7 +37,9 @@ from ..utils import (
     LazyList,
     merge_dicts,
     mimetype2ext,
+    NO_DEFAULT,
     parse_codecs,
+    parse_count,
     parse_duration,
     parse_qs,
     qualities,
@@ -44,7 +47,9 @@ from ..utils import (
     smuggle_url,
     str_or_none,
     str_to_int,
+    T,
     traverse_obj,
+    try_call,
     try_get,
     txt_or_none,
     unescapeHTML,
@@ -1247,7 +1252,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 'title': 'IMG 3456',
                 'description': '',
                 'upload_date': '20170613',
-                'uploader': 'ElevageOrVert',
+                'uploader': "l'Or Vert asbl",
                 'uploader_id': '@ElevageOrVert',
             },
             'params': {
@@ -1460,6 +1465,30 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         self._code_cache = {}
         self._player_cache = {}
 
+    # *ytcfgs, webpage=None
+    def _extract_player_url(self, *ytcfgs, **kw_webpage):
+        if ytcfgs and not isinstance(ytcfgs[0], dict):
+            webpage = kw_webpage.get('webpage') or ytcfgs[0]
+        if webpage:
+            player_url = self._search_regex(
+                r'"(?:PLAYER_JS_URL|jsUrl)"\s*:\s*"([^"]+)"',
+                webpage or '', 'player URL', fatal=False)
+            if player_url:
+                ytcfgs = ytcfgs + ({'PLAYER_JS_URL': player_url},)
+        return traverse_obj(
+            ytcfgs, (Ellipsis, 'PLAYER_JS_URL'), (Ellipsis, 'WEB_PLAYER_CONTEXT_CONFIGS', Ellipsis, 'jsUrl'),
+            get_all=False, expected_type=lambda u: urljoin('https://www.youtube.com', u))
+
+    def _download_player_url(self, video_id, fatal=False):
+        res = self._download_webpage(
+            'https://www.youtube.com/iframe_api',
+            note='Downloading iframe API JS', video_id=video_id, fatal=fatal)
+        player_version = self._search_regex(
+            r'player\\?/([0-9a-fA-F]{8})\\?/', res or '', 'player version', fatal=fatal,
+            default=NO_DEFAULT if res else None)
+        if player_version:
+            return 'https://www.youtube.com/s/player/{0}/player_ias.vflset/en_US/base.js'.format(player_version)
+
     def _signature_cache_id(self, example_sig):
         """ Return a string representation of a signature """
         return '.'.join(compat_str(len(part)) for part in example_sig.split('.'))
@@ -1474,46 +1503,49 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             raise ExtractorError('Cannot identify player %r' % player_url)
         return id_m.group('id')
 
-    def _get_player_code(self, video_id, player_url, player_id=None):
+    def _load_player(self, video_id, player_url, fatal=True, player_id=None):
         if not player_id:
             player_id = self._extract_player_info(player_url)
-
         if player_id not in self._code_cache:
-            self._code_cache[player_id] = self._download_webpage(
-                player_url, video_id,
+            code = self._download_webpage(
+                player_url, video_id, fatal=fatal,
                 note='Downloading player ' + player_id,
                 errnote='Download of %s failed' % player_url)
-        return self._code_cache[player_id]
+            if code:
+                self._code_cache[player_id] = code
+        return self._code_cache[player_id] if fatal else self._code_cache.get(player_id)
 
     def _extract_signature_function(self, video_id, player_url, example_sig):
         player_id = self._extract_player_info(player_url)
 
         # Read from filesystem cache
-        func_id = 'js_%s_%s' % (
+        func_id = 'js_{0}_{1}'.format(
             player_id, self._signature_cache_id(example_sig))
         assert os.path.basename(func_id) == func_id
 
-        cache_spec = self._downloader.cache.load('youtube-sigfuncs', func_id)
-        if cache_spec is not None:
-            return lambda s: ''.join(s[i] for i in cache_spec)
+        self.write_debug('Extracting signature function {0}'.format(func_id))
+        cache_spec, code = self.cache.load('youtube-sigfuncs', func_id), None
 
-        code = self._get_player_code(video_id, player_url, player_id)
-        res = self._parse_sig_js(code)
+        if not cache_spec:
+            code = self._load_player(video_id, player_url, player_id)
+        if code:
+            res = self._parse_sig_js(code)
+            test_string = ''.join(map(compat_chr, range(len(example_sig))))
+            cache_spec = [ord(c) for c in res(test_string)]
+            self.cache.store('youtube-sigfuncs', func_id, cache_spec)
 
-        test_string = ''.join(map(compat_chr, range(len(example_sig))))
-        cache_res = res(test_string)
-        cache_spec = [ord(c) for c in cache_res]
-
-        self._downloader.cache.store('youtube-sigfuncs', func_id, cache_spec)
-        return res
+        return lambda s: ''.join(s[i] for i in cache_spec)
 
     def _print_sig_code(self, func, example_sig):
+        if not self.get_param('youtube_print_sig_code'):
+            return
+
         def gen_sig_code(idxs):
             def _genslice(start, end, step):
                 starts = '' if start == 0 else str(start)
                 ends = (':%d' % (end + step)) if end + step >= 0 else ':'
                 steps = '' if step == 1 else (':%d' % step)
-                return 's[%s%s%s]' % (starts, ends, steps)
+                return 's[{0}{1}{2}]'.format(starts, ends, steps)
 
             step = None
             # Quelch pyflakes warnings - start will be set when step is set
@@ -1564,143 +1596,137 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             jscode, 'Initial JS player signature function name', group='sig')
 
         jsi = JSInterpreter(jscode)
-
         initial_function = jsi.extract_function(funcname)
-
         return lambda s: initial_function([s])
 
+    def _cached(self, func, *cache_id):
+        def inner(*args, **kwargs):
+            if cache_id not in self._player_cache:
+                try:
+                    self._player_cache[cache_id] = func(*args, **kwargs)
+                except ExtractorError as e:
+                    self._player_cache[cache_id] = e
+                except Exception as e:
+                    self._player_cache[cache_id] = ExtractorError(traceback.format_exc(), cause=e)
+
+            ret = self._player_cache[cache_id]
+            if isinstance(ret, Exception):
+                raise ret
+            return ret
+        return inner
+
     def _decrypt_signature(self, s, video_id, player_url):
         """Turn the encrypted s field into a working signature"""
-
-        if player_url is None:
-            raise ExtractorError('Cannot decrypt signature without player_url')
-
-        try:
-            player_id = (player_url, self._signature_cache_id(s))
-            if player_id not in self._player_cache:
-                func = self._extract_signature_function(
-                    video_id, player_url, s
-                )
-                self._player_cache[player_id] = func
-            func = self._player_cache[player_id]
-            if self._downloader.params.get('youtube_print_sig_code'):
-                self._print_sig_code(func, s)
-            return func(s)
-        except Exception as e:
-            tb = traceback.format_exc()
-            raise ExtractorError(
-                'Signature extraction failed: ' + tb, cause=e)
-
-    def _extract_player_url(self, webpage):
-        player_url = self._search_regex(
-            r'"(?:PLAYER_JS_URL|jsUrl)"\s*:\s*"([^"]+)"',
-            webpage or '', 'player URL', fatal=False)
-        if not player_url:
-            return
-        if player_url.startswith('//'):
-            player_url = 'https:' + player_url
-        elif not re.match(r'https?://', player_url):
-            player_url = compat_urllib_parse.urljoin(
-                'https://www.youtube.com', player_url)
-        return player_url
+        extract_sig = self._cached(
+            self._extract_signature_function, 'sig', player_url, self._signature_cache_id(s))
+        func = extract_sig(video_id, player_url, s)
+        self._print_sig_code(func, s)
+        return func(s)
 
     # from yt-dlp
     # See also:
     # 1. https://github.com/ytdl-org/youtube-dl/issues/29326#issuecomment-894619419
     # 2. https://code.videolan.org/videolan/vlc/-/blob/4fb284e5af69aa9ac2100ccbdd3b88debec9987f/share/lua/playlist/youtube.lua#L116
     # 3. https://github.com/ytdl-org/youtube-dl/issues/30097#issuecomment-950157377
-    def _extract_n_function_name(self, jscode):
-        target = r'(?P<nfunc>[a-zA-Z_$][\w$]*)(?:\[(?P<idx>\d+)\])?'
-        nfunc_and_idx = self._search_regex(
-            r'\.get\("n"\)\)&&\(b=(%s)\([\w$]+\)' % (target, ),
-            jscode, 'Initial JS player n function name')
-        nfunc, idx = re.match(target, nfunc_and_idx).group('nfunc', 'idx')
-        if not idx:
-            return nfunc
-
-        VAR_RE_TMPL = r'var\s+%s\s*=\s*(?P<name>\[(?P<alias>%s)\])[;,]'
-        note = 'Initial JS player n function {0} (%s[%s])' % (nfunc, idx)
-
-        def search_function_code(needle, group):
-            return self._search_regex(
-                VAR_RE_TMPL % (re.escape(nfunc), needle), jscode,
-                note.format(group), group=group)
-
-        if int_or_none(idx) == 0:
-            real_nfunc = search_function_code(r'[a-zA-Z_$][\w$]*', group='alias')
-            if real_nfunc:
-                return real_nfunc
-        return self._parse_json(
-            search_function_code('.+?', group='name'),
-            nfunc, transform_source=js_to_json)[int(idx)]
-
-    def _extract_n_function(self, video_id, player_url):
-        player_id = self._extract_player_info(player_url)
-        func_code = self._downloader.cache.load('youtube-nsig', player_id)
-
-        if func_code:
-            jsi = JSInterpreter(func_code)
-        else:
-            jscode = self._get_player_code(video_id, player_url, player_id)
-            funcname = self._extract_n_function_name(jscode)
-            jsi = JSInterpreter(jscode)
-            func_code = jsi.extract_function_code(funcname)
-            self._downloader.cache.store('youtube-nsig', player_id, func_code)
-
-        if self._downloader.params.get('youtube_print_sig_code'):
-            self.to_screen('Extracted nsig function from {0}:\n{1}\n'.format(player_id, func_code[1]))
-
-        return lambda s: jsi.extract_function_from_code(*func_code)([s])
-
-    def _n_descramble(self, n_param, player_url, video_id):
-        """Compute the response to YT's "n" parameter challenge,
-           or None
-
-        Args:
-        n_param     -- challenge string that is the value of the
-                       URL's "n" query parameter
-        player_url  -- URL of YT player JS
-        video_id
-        """
-
-        sig_id = ('nsig_value', n_param)
-        if sig_id in self._player_cache:
-            return self._player_cache[sig_id]
+    def _decrypt_nsig(self, n, video_id, player_url):
+        """Turn the encrypted n field into a working signature"""
+        if player_url is None:
+            raise ExtractorError('Cannot decrypt nsig without player_url')
 
         try:
-            player_id = ('nsig', player_url)
-            if player_id not in self._player_cache:
-                self._player_cache[player_id] = self._extract_n_function(video_id, player_url)
-            func = self._player_cache[player_id]
-            ret = func(n_param)
-            if ret.startswith('enhanced_except_'):
-                raise ExtractorError('Unhandled exception in decode')
-            self._player_cache[sig_id] = ret
-            if self._downloader.params.get('verbose', False):
-                self._downloader.to_screen('[debug] [%s] %s' % (self.IE_NAME, 'Decrypted nsig {0} => {1}'.format(n_param, self._player_cache[sig_id])))
-            return self._player_cache[sig_id]
-        except Exception as e:
-            self._downloader.report_warning(
-                '[%s] %s (%s %s)' % (
-                    self.IE_NAME,
+            jsi, player_id, func_code = self._extract_n_function_code(video_id, player_url)
+        except ExtractorError as e:
+            raise ExtractorError('Unable to extract nsig jsi, player_id, func_codefunction code', cause=e)
+        if self.get_param('youtube_print_sig_code'):
+            self.to_screen('Extracted nsig function from {0}:\n{1}\n'.format(
+                player_id, func_code[1]))
+
+        try:
+            extract_nsig = self._cached(self._extract_n_function_from_code, 'nsig func', player_url)
+            ret = extract_nsig(jsi, func_code)(n)
+        except JSInterpreter.Exception as e:
+            self.report_warning(
+                '%s (%s %s)' % (
                     'Unable to decode n-parameter: download likely to be throttled',
                     error_to_compat_str(e),
-                    traceback.format_exc()))
+                    traceback.format_exc()),
+                video_id=video_id)
+            return
+
+        self.write_debug('Decrypted nsig {0} => {1}'.format(n, ret))
+        return ret
+
+    def _extract_n_function_name(self, jscode):
+        func_name, idx = self._search_regex(
+            r'\.get\("n"\)\)&&\(b=(?P<nfunc>[a-zA-Z_$][\w$]*)(?:\[(?P<idx>\d+)\])?\([\w$]+\)',
+            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,
+            'Initial JS player n function list ({0}.{1})'.format(func_name, idx)),
+            func_name, transform_source=js_to_json)[int(idx)]
+
+    def _extract_n_function_code(self, video_id, player_url):
+        player_id = self._extract_player_info(player_url)
+        func_code = self.cache.load('youtube-nsig', player_id)
+        jscode = func_code or self._load_player(video_id, player_url)
+        jsi = JSInterpreter(jscode)
+
+        if func_code:
+            return jsi, player_id, func_code
+
+        func_name = self._extract_n_function_name(jscode)
+
+        # For redundancy
+        func_code = self._search_regex(
+            r'''(?xs)%s\s*=\s*function\s*\((?P<var>[\w$]+)\)\s*
+                     # NB: The end of the regex is intentionally kept strict
+                     {(?P<code>.+?}\s*return\ [\w$]+.join\(""\))};''' % func_name,
+            jscode, 'nsig function', group=('var', 'code'), default=None)
+        if func_code:
+            func_code = ([func_code[0]], func_code[1])
+        else:
+            self.write_debug('Extracting nsig function with jsinterp')
+            func_code = jsi.extract_function_code(func_name)
+
+        self.cache.store('youtube-nsig', player_id, func_code)
+        return jsi, player_id, func_code
+
+    def _extract_n_function_from_code(self, jsi, func_code):
+        func = jsi.extract_function_from_code(*func_code)
+
+        def extract_nsig(s):
+            try:
+                ret = func([s])
+            except JSInterpreter.Exception:
+                raise
+            except Exception as e:
+                raise JSInterpreter.Exception(traceback.format_exc(), cause=e)
+
+            if ret.startswith('enhanced_except_'):
+                raise JSInterpreter.Exception('Signature function returned an exception')
+            return ret
+
+        return extract_nsig
+
+    def _unthrottle_format_urls(self, video_id, player_url, *formats):
+
+        def decrypt_nsig(n):
+            return self._cached(self._decrypt_nsig, 'nsig', n, player_url)
 
-    def _unthrottle_format_urls(self, video_id, player_url, formats):
         for fmt in formats:
             parsed_fmt_url = compat_urllib_parse.urlparse(fmt['url'])
             n_param = compat_parse_qs(parsed_fmt_url.query).get('n')
             if not n_param:
                 continue
             n_param = n_param[-1]
-            n_response = self._n_descramble(n_param, player_url, video_id)
+            n_response = decrypt_nsig(n_param)(n_param, video_id, player_url)
             if n_response is None:
                 # give up if descrambling failed
                 break
-            for fmt_dct in traverse_obj(fmt, (None, (None, ('fragments', Ellipsis))), expected_type=dict):
-                fmt_dct['url'] = update_url(
-                    fmt_dct['url'], query_update={'n': [n_response]})
+            fmt['url'] = update_url_query(fmt['url'], {'n': n_response})
 
     # from yt-dlp, with tweaks
     def _extract_signature_timestamp(self, video_id, player_url, ytcfg=None, fatal=False):
@@ -1708,16 +1734,16 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         Extract signatureTimestamp (sts)
         Required to tell API what sig/player version is in use.
         """
-        sts = int_or_none(ytcfg.get('STS')) if isinstance(ytcfg, dict) else None
+        sts = traverse_obj(ytcfg, 'STS', expected_type=int)
         if not sts:
             # Attempt to extract from player
             if player_url is None:
                 error_msg = 'Cannot extract signature timestamp without player_url.'
                 if fatal:
                     raise ExtractorError(error_msg)
-                self._downloader.report_warning(error_msg)
+                self.report_warning(error_msg)
                 return
-            code = self._get_player_code(video_id, player_url)
+            code = self._load_player(video_id, player_url, fatal=fatal)
             sts = int_or_none(self._search_regex(
                 r'(?:signatureTimestamp|sts)\s*:\s*(?P<sts>[0-9]{5})', code or '',
                 'JS player signature timestamp', group='sts', fatal=fatal))
@@ -1733,12 +1759,18 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         # cpn generation algorithm is reverse engineered from base.js.
         # In fact it works even with dummy cpn.
         CPN_ALPHABET = 'abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789-_'
-        cpn = ''.join((CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(0, 16)))
+        cpn = ''.join(CPN_ALPHABET[random.randint(0, 256) & 63] for _ in range(0, 16))
 
-        playback_url = update_url(
-            playback_url, query_update={
-                'ver': ['2'],
-                'cpn': [cpn],
+        # more consistent results setting it to right before the end
+        qs = parse_qs(playback_url)
+        video_length = '{0}'.format(float((qs.get('len') or ['1.5'])[0]) - 1)
+
+        playback_url = update_url_query(
+            playback_url, {
+                'ver': '2',
+                'cpn': cpn,
+                'cmt': video_length,
+                'el': 'detailpage',  # otherwise defaults to "shorts"
             })
 
         self._download_webpage(
@@ -1986,8 +2018,11 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             else:
                 self.to_screen('Downloading just video %s because of --no-playlist' % video_id)
 
+        if not player_url:
+            player_url = self._extract_player_url(webpage)
+
         formats = []
-        itags = []
+        itags = collections.defaultdict(set)
         itag_qualities = {}
         q = qualities(['tiny', 'small', 'medium', 'large', 'hd720', 'hd1080', 'hd1440', 'hd2160', 'hd2880', 'highres'])
         CHUNK_SIZE = 10 << 20
@@ -2003,58 +2038,92 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                 })
             } for range_start in range(0, f['filesize'], CHUNK_SIZE))
 
+        lower = lambda s: s.lower()
+
         for fmt in streaming_formats:
-            if fmt.get('targetDurationSec') or fmt.get('drmFamilies'):
+            if fmt.get('targetDurationSec'):
                 continue
 
             itag = str_or_none(fmt.get('itag'))
-            quality = fmt.get('quality')
-            if itag and quality:
+            audio_track = traverse_obj(fmt, ('audioTrack', T(dict))) or {}
+
+            quality = traverse_obj(fmt, ((
+                # The 3gp format (17) in android client has a quality of "small",
+                # but is actually worse than other formats
+                T(lambda _: 'tiny' if itag == 17 else None),
+                ('quality', T(lambda q: q if q and q != 'tiny' else None)),
+                ('audioQuality', T(lower)),
+                'quality'), T(txt_or_none)), get_all=False)
+            if quality and itag:
                 itag_qualities[itag] = quality
             # FORMAT_STREAM_TYPE_OTF(otf=1) requires downloading the init fragment
             # (adding `&sq=0` to the URL) and parsing emsg box to determine the
-            # number of fragment that would subsequently requested with (`&sq=N`)
+            # number of fragments that would subsequently be requested with (`&sq=N`)
             if fmt.get('type') == 'FORMAT_STREAM_TYPE_OTF':
                 continue
 
             fmt_url = fmt.get('url')
             if not fmt_url:
                 sc = compat_parse_qs(fmt.get('signatureCipher'))
-                fmt_url = url_or_none(try_get(sc, lambda x: x['url'][0]))
-                encrypted_sig = try_get(sc, lambda x: x['s'][0])
-                if not (sc and fmt_url and encrypted_sig):
+                fmt_url = traverse_obj(sc, ('url', -1, T(url_or_none)))
+                encrypted_sig = traverse_obj(sc, ('s', -1))
+                if not (fmt_url and encrypted_sig):
                     continue
-                if not player_url:
-                    player_url = self._extract_player_url(webpage)
+                player_url = player_url or self._extract_player_url(webpage)
                 if not player_url:
                     continue
-                signature = self._decrypt_signature(sc['s'][0], video_id, player_url)
-                sp = try_get(sc, lambda x: x['sp'][0]) or 'signature'
-                fmt_url += '&' + sp + '=' + signature
+                try:
+                    fmt_url = update_url_query(fmt_url, {
+                        traverse_obj(sc, ('sp', -1)) or 'signature':
+                            [self._decrypt_signature(encrypted_sig, video_id, player_url)],
+                    })
+                except ExtractorError as e:
+                    self.report_warning('Signature extraction failed: Some formats may be missing',
+                                        video_id=video_id, only_once=True)
+                    self.write_debug(error_to_compat_str(e), only_once=True)
+                    continue
 
-            if itag:
-                itags.append(itag)
-            tbr = float_or_none(
-                fmt.get('averageBitrate') or fmt.get('bitrate'), 1000)
+            language_preference = (
+                10 if audio_track.get('audioIsDefault')
+                else -10 if 'descriptive' in (traverse_obj(audio_track, ('displayName', T(lower))) or '')
+                else -1)
+            name = (
+                traverse_obj(fmt, ('qualityLabel', T(txt_or_none)))
+                or quality.replace('audio_quality_', ''))
             dct = {
-                'asr': int_or_none(fmt.get('audioSampleRate')),
-                'filesize': int_or_none(fmt.get('contentLength')),
-                'format_id': itag,
-                'format_note': fmt.get('qualityLabel') or quality,
-                'fps': int_or_none(fmt.get('fps')),
-                'height': int_or_none(fmt.get('height')),
-                'quality': q(quality),
-                'tbr': tbr,
+                'format_id': join_nonempty(itag, fmt.get('isDrc') and 'drc'),
                 'url': fmt_url,
-                'width': fmt.get('width'),
+                # Format 22 is likely to be damaged: see https://github.com/yt-dlp/yt-dlp/issues/3372
+                'source_preference': ((-5 if itag == '22' else -1)
+                                      + (100 if 'Premium' in name else 0)),
+                'quality': q(quality),
+                'language': join_nonempty(audio_track.get('id', '').split('.')[0],
+                                          'desc' if language_preference < -1 else '') or None,
+                'language_preference': language_preference,
+                # Strictly de-prioritize 3gp formats
+                'preference': -2 if itag == '17' else None,
             }
-            mimetype = fmt.get('mimeType')
-            if mimetype:
-                mobj = re.match(
-                    r'((?:[^/]+)/(?:[^;]+))(?:;\s*codecs="([^"]+)")?', mimetype)
-                if mobj:
-                    dct['ext'] = mimetype2ext(mobj.group(1))
-                    dct.update(parse_codecs(mobj.group(2)))
+            if itag:
+                itags[itag].add(('https', dct.get('language')))
+            self._unthrottle_format_urls(video_id, player_url, dct)
+            dct.update(traverse_obj(fmt, {
+                'asr': ('audioSampleRate', T(int_or_none)),
+                'filesize': ('contentLength', T(int_or_none)),
+                'format_note': ('qualityLabel', T(lambda x: x or quality)),
+                # for some formats, fps is wrongly returned as 1
+                'fps': ('fps', T(int_or_none), T(lambda f: f if f > 1 else None)),
+                'audio_channels': ('audioChannels', T(int_or_none)),
+                'height': ('height', T(int_or_none)),
+                'has_drm': ('drmFamilies', T(bool)),
+                'tbr': (('averageBitrate', 'bitrate'), T(lambda t: float_or_none(t, 1000))),
+                'width': ('width', T(int_or_none)),
+                '_duration_ms': ('approxDurationMs', T(int_or_none)),
+            }, get_all=False))
+            mime_mobj = re.match(
+                r'((?:[^/]+)/(?:[^;]+))(?:;\s*codecs="([^"]+)")?', fmt.get('mimeType') or '')
+            if mime_mobj:
+                dct['ext'] = mimetype2ext(mime_mobj.group(1))
+                dct.update(parse_codecs(mime_mobj.group(2)))
             single_stream = 'none' in (dct.get(c) for c in ('acodec', 'vcodec'))
             if single_stream and dct.get('ext'):
                 dct['container'] = dct['ext'] + '_dash'
@@ -2069,32 +2138,62 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
 
             formats.append(dct)
 
+        def process_manifest_format(f, proto, client_name, itag, all_formats=False):
+            key = (proto, f.get('language'))
+            if not all_formats and key in itags[itag]:
+                return False
+            itags[itag].add(key)
+
+            if itag:
+                f['format_id'] = (
+                    '{0}-{1}'.format(itag, proto)
+                    if all_formats or any(p != proto for p, _ in itags[itag])
+                    else itag)
+
+            if f.get('source_preference') is None:
+                f['source_preference'] = -1
+
+            if itag in ('616', '235'):
+                f['format_note'] = join_nonempty(f.get('format_note'), 'Premium', delim=' ')
+                f['source_preference'] += 100
+
+            f['quality'] = q(traverse_obj(f, (
+                'format_id', T(lambda s: itag_qualities[s.split('-')[0]])), default=-1))
+            if try_call(lambda: f['fps'] <= 1):
+                del f['fps']
+
+            if proto == 'hls' and f.get('has_drm'):
+                f['has_drm'] = 'maybe'
+                f['source_preference'] -= 5
+            return True
+
         hls_manifest_url = streaming_data.get('hlsManifestUrl')
         if hls_manifest_url:
             for f in self._extract_m3u8_formats(
                     hls_manifest_url, video_id, 'mp4', fatal=False):
-                itag = self._search_regex(
-                    r'/itag/(\d+)', f['url'], 'itag', default=None)
-                if itag:
-                    f['format_id'] = itag
-                formats.append(f)
+                if process_manifest_format(
+                        f, 'hls', None, self._search_regex(
+                            r'/itag/(\d+)', f['url'], 'itag', default=None)):
+                    formats.append(f)
 
         if self._downloader.params.get('youtube_include_dash_manifest', True):
             dash_manifest_url = streaming_data.get('dashManifestUrl')
             if dash_manifest_url:
                 for f in self._extract_mpd_formats(
                         dash_manifest_url, video_id, fatal=False):
-                    itag = f['format_id']
-                    if itag in itags:
-                        continue
-                    if itag in itag_qualities:
-                        f['quality'] = q(itag_qualities[itag])
-                    filesize = int_or_none(self._search_regex(
-                        r'/clen/(\d+)', f.get('fragment_base_url')
-                        or f['url'], 'file size', default=None))
-                    if filesize:
-                        f['filesize'] = filesize
-                    formats.append(f)
+                    if process_manifest_format(
+                            f, 'dash', None, f['format_id']):
+                        f['filesize'] = traverse_obj(f, (
+                            ('fragment_base_url', 'url'), T(lambda u: self._search_regex(
+                                r'/clen/(\d+)', u, 'file size', default=None)),
+                            T(int_or_none)), get_all=False)
+                        formats.append(f)
+
+        playable_formats = [f for f in formats if not f.get('has_drm')]
+        if formats and not playable_formats:
+            # If there are no formats that definitely don't have DRM, all have DRM
+            self.report_drm(video_id)
+        formats[:] = playable_formats
 
         if not formats:
             if streaming_data.get('licenseInfos'):
@@ -2166,6 +2265,17 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
             video_details.get('lengthSeconds')
             or microformat.get('lengthSeconds')) \
             or parse_duration(search_meta('duration'))
+
+        for f in formats:
+            # Some formats may have much smaller duration than others (possibly damaged during encoding)
+            # but avoid false positives with small duration differences.
+            # Ref: https://github.com/yt-dlp/yt-dlp/issues/2823
+            if try_call(lambda x: float(x.pop('_duration_ms')) / duration < 500, args=(f,)):
+                self.report_warning(
+                    '{0}: Some possibly damaged formats will be deprioritized'.format(video_id), only_once=True)
+                # Strictly de-prioritize damaged formats
+                f['preference'] = -10
+
         is_live = video_details.get('isLive')
 
         owner_profile_url = self._yt_urljoin(self._extract_author_var(
@@ -2174,10 +2284,6 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
         uploader = self._extract_author_var(
             webpage, 'name', videodetails=video_details, metadata=microformat)
 
-        if not player_url:
-            player_url = self._extract_player_url(webpage)
-        self._unthrottle_format_urls(video_id, player_url, formats)
-
         info = {
             'id': video_id,
             'title': self._live_title(video_title) if is_live else video_title,
@@ -2370,6 +2476,14 @@ class YoutubeIE(YoutubeBaseInfoExtractor):
                             'like_count': str_to_int(like_count),
                             'dislike_count': str_to_int(dislike_count),
                         })
+                    else:
+                        info['like_count'] = traverse_obj(vpir, (
+                            'videoActions', 'menuRenderer', 'topLevelButtons', Ellipsis,
+                            'segmentedLikeDislikeButtonViewModel', 'likeButtonViewModel', 'likeButtonViewModel',
+                            'toggleButtonViewModel', 'toggleButtonViewModel', 'defaultButtonViewModel',
+                            'buttonViewModel', (('title', ('accessibilityText', T(lambda s: s.split()), Ellipsis))), T(parse_count)),
+                            get_all=False)
+
                 vsir = content.get('videoSecondaryInfoRenderer')
                 if vsir:
                     rows = try_get(
@@ -2484,7 +2598,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
         'playlist_mincount': 94,
         'info_dict': {
             'id': 'UCqj7Cz7revf5maW9g5pgNcg',
-            'title': 'Igor Kleiner - Playlists',
+            'title': r're:Igor Kleiner(?: Ph\.D\.)? - Playlists',
             'description': 'md5:be97ee0f14ee314f1f002cf187166ee2',
             'uploader': 'Igor Kleiner',
             'uploader_id': '@IgorDataScience',
@@ -2495,7 +2609,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
         'playlist_mincount': 94,
         'info_dict': {
             'id': 'UCqj7Cz7revf5maW9g5pgNcg',
-            'title': 'Igor Kleiner - Playlists',
+            'title': r're:Igor Kleiner(?: Ph\.D\.)? - Playlists',
             'description': 'md5:be97ee0f14ee314f1f002cf187166ee2',
             'uploader': 'Igor Kleiner',
             'uploader_id': '@IgorDataScience',
@@ -2607,7 +2721,7 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor):
         'url': 'https://www.youtube.com/channel/UCKfVa3S1e4PHvxWcwyMMg8w/channels',
         'info_dict': {
             'id': 'UCKfVa3S1e4PHvxWcwyMMg8w',
-            'title': 'lex will - Channels',
+            'title': r're:lex will - (?:Home|Channels)',
             'description': 'md5:2163c5d0ff54ed5f598d6a7e6211e488',
             'uploader': 'lex will',
             'uploader_id': '@lexwill718',

youtube_dl/postprocessor/ffmpeg.py

@@ -74,8 +74,11 @@ class FFmpegPostProcessor(PostProcessor):
         return FFmpegPostProcessor(downloader)._versions
 
     def _determine_executables(self):
-        programs = ['avprobe', 'avconv', 'ffmpeg', 'ffprobe']
+        # ordered to match prefer_ffmpeg!
+        convs = ['ffmpeg', 'avconv']
+        probes = ['ffprobe', 'avprobe']
         prefer_ffmpeg = True
+        programs = convs + probes
 
         def get_ffmpeg_version(path):
             ver = get_exe_version(path, args=['-version'])
@@ -96,6 +99,7 @@ class FFmpegPostProcessor(PostProcessor):
 
         self._paths = None
         self._versions = None
+        location = None
         if self._downloader:
             prefer_ffmpeg = self._downloader.params.get('prefer_ffmpeg', True)
             location = self._downloader.params.get('ffmpeg_location')
@@ -118,33 +122,21 @@ class FFmpegPostProcessor(PostProcessor):
                     location = os.path.dirname(os.path.abspath(location))
                     if basename in ('ffmpeg', 'ffprobe'):
                         prefer_ffmpeg = True
+        self._paths = dict(
+            (p, p if location is None else os.path.join(location, p))
+            for p in programs)
+        self._versions = dict(
+            x for x in (
+                (p, get_ffmpeg_version(self._paths[p])) for p in programs)
+            if x[1] is not None)
 
-                self._paths = dict(
-                    (p, os.path.join(location, p)) for p in programs)
-                self._versions = dict(
-                    (p, get_ffmpeg_version(self._paths[p])) for p in programs)
-        if self._versions is None:
-            self._versions = dict(
-                (p, get_ffmpeg_version(p)) for p in programs)
-            self._paths = dict((p, p) for p in programs)
-
-        if prefer_ffmpeg is False:
-            prefs = ('avconv', 'ffmpeg')
-        else:
-            prefs = ('ffmpeg', 'avconv')
-        for p in prefs:
-            if self._versions[p]:
-                self.basename = p
-                break
-
-        if prefer_ffmpeg is False:
-            prefs = ('avprobe', 'ffprobe')
-        else:
-            prefs = ('ffprobe', 'avprobe')
-        for p in prefs:
-            if self._versions[p]:
-                self.probe_basename = p
-                break
+        basenames = [None, None]
+        for i, progs in enumerate((convs, probes)):
+            for p in progs[::-1 if prefer_ffmpeg is False else 1]:
+                if self._versions.get(p):
+                    basenames[i] = p
+                    break
+        self.basename, self.probe_basename = basenames
 
     @property
     def available(self):

youtube_dl/utils.py

@@ -45,6 +45,7 @@ from .compat import (
     compat_casefold,
     compat_chr,
     compat_collections_abc,
+    compat_contextlib_suppress,
     compat_cookiejar,
     compat_ctypes_WINFUNCTYPE,
     compat_datetime_timedelta_total_seconds,
@@ -1855,25 +1856,18 @@ def write_json_file(obj, fn):
     try:
         with tf:
             json.dump(obj, tf)
-        if sys.platform == 'win32':
-            # Need to remove existing file on Windows, else os.rename raises
-            # WindowsError or FileExistsError.
-            try:
+        with compat_contextlib_suppress(OSError):
+            if sys.platform == 'win32':
+                # Need to remove existing file on Windows, else os.rename raises
+                # WindowsError or FileExistsError.
                 os.unlink(fn)
-            except OSError:
-                pass
-        try:
             mask = os.umask(0)
             os.umask(mask)
             os.chmod(tf.name, 0o666 & ~mask)
-        except OSError:
-            pass
         os.rename(tf.name, fn)
     except Exception:
-        try:
+        with compat_contextlib_suppress(OSError):
             os.remove(tf.name)
-        except OSError:
-            pass
         raise
 
 
@@ -2033,14 +2027,13 @@ def extract_attributes(html_element):
     NB HTMLParser is stricter in Python 2.6 & 3.2 than in later versions,
     but the cases in the unit test will work for all of 2.6, 2.7, 3.2-3.5.
     """
-    parser = HTMLAttributeParser()
-    try:
-        parser.feed(html_element)
-        parser.close()
-    # Older Python may throw HTMLParseError in case of malformed HTML
-    except compat_HTMLParseError:
-        pass
-    return parser.attrs
+    ret = None
+    # Older Python may throw HTMLParseError in case of malformed HTML (and on .close()!)
+    with compat_contextlib_suppress(compat_HTMLParseError):
+        with contextlib.closing(HTMLAttributeParser()) as parser:
+            parser.feed(html_element)
+            ret = parser.attrs
+    return ret or {}
 
 
 def clean_html(html):
@@ -2182,8 +2175,28 @@ def sanitize_url(url):
     return url
 
 
+def extract_basic_auth(url):
+    parts = compat_urllib_parse.urlsplit(url)
+    if parts.username is None:
+        return url, None
+    url = compat_urllib_parse.urlunsplit(parts._replace(netloc=(
+        parts.hostname if parts.port is None
+        else '%s:%d' % (parts.hostname, parts.port))))
+    auth_payload = base64.b64encode(
+        ('%s:%s' % (parts.username, parts.password or '')).encode('utf-8'))
+    return url, 'Basic {0}'.format(auth_payload.decode('ascii'))
+
+
 def sanitized_Request(url, *args, **kwargs):
-    return compat_urllib_request.Request(escape_url(sanitize_url(url)), *args, **kwargs)
+    url, auth_header = extract_basic_auth(escape_url(sanitize_url(url)))
+    if auth_header is not None:
+        headers = args[1] if len(args) > 1 else kwargs.get('headers')
+        headers = headers or {}
+        headers['Authorization'] = auth_header
+        if len(args) <= 1 and kwargs.get('headers') is None:
+            kwargs['headers'] = headers
+            kwargs = compat_kwargs(kwargs)
+    return compat_urllib_request.Request(url, *args, **kwargs)
 
 
 def expand_path(s):
@@ -2221,7 +2234,8 @@ def _htmlentity_transform(entity_with_semicolon):
             numstr = '0%s' % numstr
         else:
             base = 10
-        # See https://github.com/ytdl-org/youtube-dl/issues/7518
+        # See https://github.com/ytdl-org/youtube-dl/issues/7518\
+        # Also, weirdly, compat_contextlib_suppress fails here in 2.6
         try:
             return compat_chr(int(numstr, base))
         except ValueError:
@@ -2328,11 +2342,9 @@ def make_HTTPS_handler(params, **kwargs):
         # Some servers may (wrongly) reject requests if ALPN extension is not sent. See:
         # https://github.com/python/cpython/issues/85140
         # https://github.com/yt-dlp/yt-dlp/issues/3878
-        try:
+        with compat_contextlib_suppress(AttributeError, NotImplementedError):
+            # fails for Python < 2.7.10, not ssl.HAS_ALPN
             ctx.set_alpn_protocols(ALPN_PROTOCOLS)
-        except (AttributeError, NotImplementedError):
-            # Python < 2.7.10, not ssl.HAS_ALPN
-            pass
 
     opts_no_check_certificate = params.get('nocheckcertificate', False)
     if hasattr(ssl, 'create_default_context'):  # Python >= 3.4 or 2.7.9
@@ -2342,12 +2354,10 @@ def make_HTTPS_handler(params, **kwargs):
             context.check_hostname = False
             context.verify_mode = ssl.CERT_NONE
 
-        try:
+        with compat_contextlib_suppress(TypeError):
+            # Fails with Python 2.7.8 (create_default_context present
+            # but HTTPSHandler has no context=)
             return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
-        except TypeError:
-            # Python 2.7.8
-            # (create_default_context present but HTTPSHandler has no context=)
-            pass
 
     if sys.version_info < (3, 2):
         return YoutubeDLHTTPSHandler(params, **kwargs)
@@ -2361,15 +2371,24 @@ def make_HTTPS_handler(params, **kwargs):
         return YoutubeDLHTTPSHandler(params, context=context, **kwargs)
 
 
-def bug_reports_message():
+def bug_reports_message(before=';'):
     if ytdl_is_updateable():
         update_cmd = 'type  youtube-dl -U  to update'
     else:
-        update_cmd = 'see  https://yt-dl.org/update  on how to update'
-    msg = '; please report this issue on https://yt-dl.org/bug .'
-    msg += ' Make sure you are using the latest version; %s.' % update_cmd
-    msg += ' Be sure to call youtube-dl with the --verbose flag and include its complete output.'
-    return msg
+        update_cmd = 'see  https://github.com/ytdl-org/youtube-dl/#user-content-installation  on how to update'
+
+    msg = (
+        'please report this issue on https://github.com/ytdl-org/youtube-dl/issues ,'
+        ' using the appropriate issue template.'
+        ' Make sure you are using the latest version; %s.'
+        ' Be sure to call youtube-dl with the --verbose option and include the complete output.'
+    ) % update_cmd
+
+    before = (before or '').rstrip()
+    if not before or before.endswith(('.', '!', '?')):
+        msg = msg[0].title() + msg[1:]
+
+    return (before + ' ' if before else '') + msg
 
 
 class YoutubeDLError(Exception):
@@ -3156,12 +3175,10 @@ def parse_iso8601(date_str, delimiter='T', timezone=None):
     if timezone is None:
         timezone, date_str = extract_timezone(date_str)
 
-    try:
+    with compat_contextlib_suppress(ValueError):
         date_format = '%Y-%m-%d{0}%H:%M:%S'.format(delimiter)
         dt = datetime.datetime.strptime(date_str, date_format) - timezone
         return calendar.timegm(dt.timetuple())
-    except ValueError:
-        pass
 
 
 def date_formats(day_first=True):
@@ -3181,17 +3198,13 @@ def unified_strdate(date_str, day_first=True):
     _, date_str = extract_timezone(date_str)
 
     for expression in date_formats(day_first):
-        try:
+        with compat_contextlib_suppress(ValueError):
             upload_date = datetime.datetime.strptime(date_str, expression).strftime('%Y%m%d')
-        except ValueError:
-            pass
     if upload_date is None:
         timetuple = email.utils.parsedate_tz(date_str)
         if timetuple:
-            try:
+            with compat_contextlib_suppress(ValueError):
                 upload_date = datetime.datetime(*timetuple[:6]).strftime('%Y%m%d')
-            except ValueError:
-                pass
     if upload_date is not None:
         return compat_str(upload_date)
 
@@ -3220,11 +3233,9 @@ def unified_timestamp(date_str, day_first=True):
         date_str = m.group(1)
 
     for expression in date_formats(day_first):
-        try:
+        with compat_contextlib_suppress(ValueError):
             dt = datetime.datetime.strptime(date_str, expression) - timezone + datetime.timedelta(hours=pm_delta)
             return calendar.timegm(dt.timetuple())
-        except ValueError:
-            pass
     timetuple = email.utils.parsedate_tz(date_str)
     if timetuple:
         return calendar.timegm(timetuple) + pm_delta * 3600 - compat_datetime_timedelta_total_seconds(timezone)
@@ -3832,14 +3843,15 @@ class PUTRequest(compat_urllib_request.Request):
         return 'PUT'
 
 
-def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1):
+def int_or_none(v, scale=1, default=None, get_attr=None, invscale=1, base=None):
     if get_attr:
         if v is not None:
             v = getattr(v, get_attr, None)
     if v in (None, ''):
         return default
     try:
-        return int(v) * invscale // scale
+        # like int, raise if base is specified and v is not a string
+        return (int(v) if base is None else int(v, base=base)) * invscale // scale
     except (ValueError, TypeError, OverflowError):
         return default