Merge c8b62b28d6 into 0153b387e5

* [VidLii] Add HD support  (yt-dlp backport-ish)

* Also fix a bug with the view count

---------

Co-authored-by: dirkf <fieldhouse@gmx.net>

test/helper.py

@@ -5,9 +5,9 @@ import hashlib
 import json
 import os.path
 import re
-import types
 import ssl
 import sys
+import types
 import unittest
 
 import youtube_dl.extractor
@@ -181,18 +181,18 @@ def expect_value(self, got, expected, field):
             op, _, expected_num = expected.partition(':')
             expected_num = int(expected_num)
             if op == 'mincount':
-                assert_func = assertGreaterEqual
+                assert_func = self.assertGreaterEqual
                 msg_tmpl = 'Expected %d items in field %s, but only got %d'
             elif op == 'maxcount':
-                assert_func = assertLessEqual
+                assert_func = self.assertLessEqual
                 msg_tmpl = 'Expected maximum %d items in field %s, but got %d'
             elif op == 'count':
-                assert_func = assertEqual
+                assert_func = self.assertEqual
                 msg_tmpl = 'Expected exactly %d items in field %s, but got %d'
             else:
                 assert False
             assert_func(
-                self, len(got), expected_num,
+                len(got), expected_num,
                 msg_tmpl % (expected_num, field, len(got)))
             return
         self.assertEqual(
@@ -262,27 +262,6 @@ def assertRegexpMatches(self, text, regexp, msg=None):
             self.assertTrue(m, msg)
 
 
-def assertGreaterEqual(self, got, expected, msg=None):
-    if not (got >= expected):
-        if msg is None:
-            msg = '%r not greater than or equal to %r' % (got, expected)
-        self.assertTrue(got >= expected, msg)
-
-
-def assertLessEqual(self, got, expected, msg=None):
-    if not (got <= expected):
-        if msg is None:
-            msg = '%r not less than or equal to %r' % (got, expected)
-        self.assertTrue(got <= expected, msg)
-
-
-def assertEqual(self, got, expected, msg=None):
-    if not (got == expected):
-        if msg is None:
-            msg = '%r not equal to %r' % (got, expected)
-        self.assertTrue(got == expected, msg)
-
-
 def expect_warnings(ydl, warnings_re):
     real_warning = ydl.report_warning
 

test/test_download.py

@@ -9,8 +9,6 @@ import unittest
 sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
 
 from test.helper import (
-    assertGreaterEqual,
-    assertLessEqual,
     expect_warnings,
     get_params,
     gettestcases,
@@ -36,12 +34,20 @@ from youtube_dl.utils import (
     ExtractorError,
     error_to_compat_str,
     format_bytes,
+    IDENTITY,
+    preferredencoding,
     UnavailableVideoError,
 )
 from youtube_dl.extractor import get_info_extractor
 
 RETRIES = 3
 
+# Some unittest APIs require actual str
+if not isinstance('TEST', str):
+    _encode_str = lambda s: s.encode(preferredencoding())
+else:
+    _encode_str = IDENTITY
+
 
 class YoutubeDL(youtube_dl.YoutubeDL):
     def __init__(self, *args, **kwargs):
@@ -102,7 +108,7 @@ def generator(test_case, tname):
 
         def print_skipping(reason):
             print('Skipping %s: %s' % (test_case['name'], reason))
-            self.skipTest(reason)
+            self.skipTest(_encode_str(reason))
 
         if not ie.working():
             print_skipping('IE marked as not _WORKING')
@@ -187,16 +193,14 @@ def generator(test_case, tname):
                 expect_info_dict(self, res_dict, test_case.get('info_dict', {}))
 
             if 'playlist_mincount' in test_case:
-                assertGreaterEqual(
-                    self,
+                self.assertGreaterEqual(
                     len(res_dict['entries']),
                     test_case['playlist_mincount'],
                     'Expected at least %d in playlist %s, but got only %d' % (
                         test_case['playlist_mincount'], test_case['url'],
                         len(res_dict['entries'])))
             if 'playlist_maxcount' in test_case:
-                assertLessEqual(
-                    self,
+                self.assertLessEqual(
                     len(res_dict['entries']),
                     test_case['playlist_maxcount'],
                     'Expected at most %d in playlist %s, but got %d' % (
@@ -243,8 +247,8 @@ def generator(test_case, tname):
                         if params.get('test'):
                             expected_minsize = max(expected_minsize, 10000)
                         got_fsize = os.path.getsize(tc_filename)
-                        assertGreaterEqual(
-                            self, got_fsize, expected_minsize,
+                        self.assertGreaterEqual(
+                            got_fsize, expected_minsize,
                             'Expected %s to be at least %s, but it\'s only %s ' %
                             (tc_filename, format_bytes(expected_minsize),
                                 format_bytes(got_fsize)))

youtube_dl/YoutubeDL.py

@@ -1039,8 +1039,8 @@ class YoutubeDL(object):
         elif result_type in ('playlist', 'multi_video'):
             # Protect from infinite recursion due to recursively nested playlists
             # (see https://github.com/ytdl-org/youtube-dl/issues/27833)
-            webpage_url = ie_result['webpage_url']
-            if webpage_url in self._playlist_urls:
+            webpage_url = ie_result.get('webpage_url')  # not all pl/mv have this
+            if webpage_url and webpage_url in self._playlist_urls:
                 self.to_screen(
                     '[download] Skipping already downloaded playlist: %s'
                     % ie_result.get('title') or ie_result.get('id'))
@@ -1048,6 +1048,10 @@ class YoutubeDL(object):
 
             self._playlist_level += 1
             self._playlist_urls.add(webpage_url)
+            new_result = dict((k, v) for k, v in extra_info.items() if k not in ie_result)
+            if new_result:
+                new_result.update(ie_result)
+                ie_result = new_result
             try:
                 return self.__process_playlist(ie_result, download)
             finally:
@@ -1593,6 +1597,28 @@ class YoutubeDL(object):
         self.cookiejar.add_cookie_header(pr)
         return pr.get_header('Cookie')
 
+    def _fill_common_fields(self, info_dict, final=True):
+
+        for ts_key, date_key in (
+                ('timestamp', 'upload_date'),
+                ('release_timestamp', 'release_date'),
+        ):
+            if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
+                # Working around out-of-range timestamp values (e.g. negative ones on Windows,
+                # see http://bugs.python.org/issue1646728)
+                try:
+                    upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
+                    info_dict[date_key] = compat_str(upload_date.strftime('%Y%m%d'))
+                except (ValueError, OverflowError, OSError):
+                    pass
+
+        # Auto generate title fields corresponding to the *_number fields when missing
+        # in order to always have clean titles. This is very common for TV series.
+        if final:
+            for field in ('chapter', 'season', 'episode'):
+                if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
+                    info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
+
     def process_video_result(self, info_dict, download=True):
         assert info_dict.get('_type', 'video') == 'video'
 
@@ -1660,24 +1686,7 @@ class YoutubeDL(object):
         if 'display_id' not in info_dict and 'id' in info_dict:
             info_dict['display_id'] = info_dict['id']
 
-        for ts_key, date_key in (
-                ('timestamp', 'upload_date'),
-                ('release_timestamp', 'release_date'),
-        ):
-            if info_dict.get(date_key) is None and info_dict.get(ts_key) is not None:
-                # Working around out-of-range timestamp values (e.g. negative ones on Windows,
-                # see http://bugs.python.org/issue1646728)
-                try:
-                    upload_date = datetime.datetime.utcfromtimestamp(info_dict[ts_key])
-                    info_dict[date_key] = compat_str(upload_date.strftime('%Y%m%d'))
-                except (ValueError, OverflowError, OSError):
-                    pass
-
-        # Auto generate title fields corresponding to the *_number fields when missing
-        # in order to always have clean titles. This is very common for TV series.
-        for field in ('chapter', 'season', 'episode'):
-            if info_dict.get('%s_number' % field) is not None and not info_dict.get(field):
-                info_dict[field] = '%s %d' % (field.capitalize(), info_dict['%s_number' % field])
+        self._fill_common_fields(info_dict)
 
         for cc_kind in ('subtitles', 'automatic_captions'):
             cc = info_dict.get(cc_kind)

youtube_dl/extractor/extractors.py

@@ -616,6 +616,7 @@ from .lego import LEGOIE
 from .lemonde import LemondeIE
 from .lenta import LentaIE
 from .libraryofcongress import LibraryOfCongressIE
+from .librivox import LibriVoxIE
 from .libsyn import LibsynIE
 from .lifenews import (
     LifeNewsIE,
@@ -898,21 +899,13 @@ from .ooyala import (
 )
 from .ora import OraTVIE
 from .orf import (
-    ORFTVthekIE,
-    ORFFM4IE,
+    ORFONIE,
+    ORFONLiveIE,
     ORFFM4StoryIE,
-    ORFOE1IE,
-    ORFOE3IE,
-    ORFNOEIE,
-    ORFWIEIE,
-    ORFBGLIE,
-    ORFOOEIE,
-    ORFSTMIE,
-    ORFKTNIE,
-    ORFSBGIE,
-    ORFTIRIE,
-    ORFVBGIE,
     ORFIPTVIE,
+    ORFPodcastIE,
+    ORFRadioIE,
+    ORFRadioCollectionIE,
 )
 from .outsidetv import OutsideTVIE
 from .packtpub import (

youtube_dl/extractor/librivox.py

@@ -0,0 +1,54 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..utils import orderedSet
+
+
+class LibriVoxIE(InfoExtractor):
+    _VALID_URL = r'https?://(?:www\.)?librivox\.org/(?P<id>(?P<title>(?:[^\-]*\-)+[^\-]*)\-by\-(?P<author>(-.*\-)*[^/]*))/?'
+    _TESTS = [{
+        'url': 'https://librivox.org/the-art-of-war-by-sun-tzu/',
+        'info_dict': {
+            'id': 'the-art-of-war-by-sun-tzu',
+            'title': 'The Art Of War by Sun Tzu',
+            'description': '"The Art of War is a Chinese military treatise written during the 6th century BC by Sun Tzu. Composed of 13 chapters, each of which is devoted to one aspect of warfare, it has long been praised as the definitive work on military strategies and tactics of its time. The Art of War is one of the oldest and most famous studies of strategy and has had a huge influence on both military planning and beyond. The Art of War has also been applied, with much success, to business and managerial strategies." (summary from Wikipedia)'
+        },
+        'playlist_mincount': 7
+    }, {
+        'url': 'https://librivox.org/alexander-the-great-by-jacob-abbott/',
+        'info_dict': {
+            'id': 'alexander-the-great-by-jacob-abbott',
+            'title': 'Alexander The Great by Jacob Abbott',
+            'description': 'Alexander the Great was one of the most successful military commanders in history, and was undefeated in battle. By the time of his death, he had conquered most of the world known to the ancient Greeks.\nAlexander the Great is one of many biographies aimed at young people written by Jacob Abbott and his brother. The biographies are written in such a way that makes them appealing and easily accessible to everyone. - Written by Wikipedia and Lizzie Driver'
+        },
+        'playlist_mincount': 12
+    }]
+
+    def _real_extract(self, url):
+        mobj = re.match(self._VALID_URL, url)
+
+        video_id = mobj.group('id')
+
+        webpage = self._download_webpage(url, video_id)
+
+        book_title = self._html_search_regex(
+            r'<h1>(?P<title>.+)</h1>', webpage, 'title', group='title', fatal=False).lower().title() or mobj.group('title').replace('-', ' ').strip().title()
+        author = mobj.group('author').replace('-', ' ').strip().title()
+        description = self._html_search_regex(
+            r'<div class=(["\'])description\1(^>)*>(<p(^>)*>)?(?P<desc>.+)(</p>)?</div>',
+            webpage, 'description', group='desc', fatal=False) or self._og_search_description(webpage)
+
+        info = {
+            'id': video_id,
+            '_type': 'playlist',
+            'title': book_title + ' by ' + author,
+            'description': description
+        }
+
+        links = orderedSet(re.findall(r'<a href=(["\'])(https?://(?:www\.)?archive\.org/download/[^/]*/([^\.]*(?<!(?:64kb)))\.mp3)\1.*>(.*)</a>', webpage))
+        info['entries'] = [self.url_result(link[1], video_id=link[2], video_title=link[3]) for link in links]
+
+        return info

youtube_dl/extractor/mixcloud.py

@@ -1,3 +1,4 @@
+# coding: utf-8
 from __future__ import unicode_literals
 
 import itertools
@@ -10,7 +11,7 @@ from ..compat import (
     compat_ord,
     compat_str,
     compat_urllib_parse_unquote,
-    compat_zip
+    compat_zip as zip,
 )
 from ..utils import (
     int_or_none,
@@ -24,7 +25,7 @@ class MixcloudBaseIE(InfoExtractor):
     def _call_api(self, object_type, object_fields, display_id, username, slug=None):
         lookup_key = object_type + 'Lookup'
         return self._download_json(
-            'https://www.mixcloud.com/graphql', display_id, query={
+            'https://app.mixcloud.com/graphql', display_id, query={
                 'query': '''{
   %s(lookup: {username: "%s"%s}) {
     %s
@@ -44,7 +45,7 @@ class MixcloudIE(MixcloudBaseIE):
             'ext': 'm4a',
             'title': 'Cryptkeeper',
             'description': 'After quite a long silence from myself, finally another Drum\'n\'Bass mix with my favourite current dance floor bangers.',
-            'uploader': 'Daniel Holbach',
+            'uploader': 'dholbach',  # was: 'Daniel Holbach',
             'uploader_id': 'dholbach',
             'thumbnail': r're:https?://.*\.jpg',
             'view_count': int,
@@ -57,7 +58,7 @@ class MixcloudIE(MixcloudBaseIE):
             'id': 'gillespeterson_caribou-7-inch-vinyl-mix-chat',
             'ext': 'mp3',
             'title': 'Caribou 7 inch Vinyl Mix & Chat',
-            'description': 'md5:2b8aec6adce69f9d41724647c65875e8',
+            'description': r're:Last week Dan Snaith aka Caribou swung by the Brownswood.{136}',
             'uploader': 'Gilles Peterson Worldwide',
             'uploader_id': 'gillespeterson',
             'thumbnail': 're:https?://.*',
@@ -65,6 +66,23 @@ class MixcloudIE(MixcloudBaseIE):
             'timestamp': 1422987057,
             'upload_date': '20150203',
         },
+        'params': {
+            'skip_download': '404 not found',
+        },
+    }, {
+        'url': 'https://www.mixcloud.com/gillespeterson/carnival-m%C3%BAsica-popular-brasileira-mix/',
+        'info_dict': {
+            'id': 'gillespeterson_carnival-música-popular-brasileira-mix',
+            'ext': 'm4a',
+            'title': 'Carnival Música Popular Brasileira Mix',
+            'description': r're:Gilles was recently in Brazil to play at Boiler Room.{208}',
+            'timestamp': 1454347174,
+            'upload_date': '20160201',
+            'uploader': 'Gilles Peterson Worldwide',
+            'uploader_id': 'gillespeterson',
+            'thumbnail': 're:https?://.*',
+            'view_count': int,
+        },
     }, {
         'url': 'https://beta.mixcloud.com/RedLightRadio/nosedrip-15-red-light-radio-01-18-2016/',
         'only_matching': True,
@@ -76,10 +94,10 @@ class MixcloudIE(MixcloudBaseIE):
         """Encrypt/Decrypt XOR cipher. Both ways are possible because it's XOR."""
         return ''.join([
             compat_chr(compat_ord(ch) ^ compat_ord(k))
-            for ch, k in compat_zip(ciphertext, itertools.cycle(key))])
+            for ch, k in zip(ciphertext, itertools.cycle(key))])
 
     def _real_extract(self, url):
-        username, slug = re.match(self._VALID_URL, url).groups()
+        username, slug = self._match_valid_url(url).groups()
         username, slug = compat_urllib_parse_unquote(username), compat_urllib_parse_unquote(slug)
         track_id = '%s_%s' % (username, slug)
 

youtube_dl/extractor/vidlii.py

@@ -4,6 +4,7 @@ from __future__ import unicode_literals
 import re
 
 from .common import InfoExtractor
+
 from ..utils import (
     float_or_none,
     get_element_by_id,
@@ -11,6 +12,7 @@ from ..utils import (
     strip_or_none,
     unified_strdate,
     urljoin,
+    str_to_int,
 )
 
 
@@ -35,6 +37,26 @@ class VidLiiIE(InfoExtractor):
             'categories': ['News & Politics'],
             'tags': ['Vidlii', 'Jan', 'Videogames'],
         }
+    }, {
+        # HD
+        'url': 'https://www.vidlii.com/watch?v=2Ng8Abj2Fkl',
+        'md5': '450e7da379c884788c3a4fa02a3ce1a4',
+        'info_dict': {
+            'id': '2Ng8Abj2Fkl',
+            'ext': 'mp4',
+            'title': 'test',
+            'description': 'md5:cc55a86032a7b6b3cbfd0f6b155b52e9',
+            'thumbnail': 'https://www.vidlii.com/usfi/thmp/2Ng8Abj2Fkl.jpg',
+            'uploader': 'VidLii',
+            'uploader_url': 'https://www.vidlii.com/user/VidLii',
+            'upload_date': '20200927',
+            'duration': 5,
+            'view_count': int,
+            'comment_count': int,
+            'average_rating': float,
+            'categories': ['Film & Animation'],
+            'tags': list,
+        },
     }, {
         'url': 'https://www.vidlii.com/embed?v=tJluaH4BJ3v&a=0',
         'only_matching': True,
@@ -46,11 +68,32 @@ class VidLiiIE(InfoExtractor):
         webpage = self._download_webpage(
             'https://www.vidlii.com/watch?v=%s' % video_id, video_id)
 
-        video_url = self._search_regex(
-            r'src\s*:\s*(["\'])(?P<url>(?:https?://)?(?:(?!\1).)+)\1', webpage,
-            'video url', group='url')
+        formats = []
 
-        title = self._search_regex(
+        def add_format(format_url, height=None):
+            height = int(self._search_regex(r'(\d+)\.mp4',
+                         format_url, 'height', default=360))
+
+            formats.append({
+                'url': format_url,
+                'format_id': '%dp' % height if height else None,
+                'height': height,
+            })
+
+        sources = re.findall(
+            r'src\s*:\s*(["\'])(?P<url>(?:https?://)?(?:(?!\1).)+)\1',
+            webpage)
+
+        formats = []
+        if len(sources) > 1:
+            add_format(sources[1][1])
+            self._check_formats(formats, video_id)
+        if len(sources) > 0:
+            add_format(sources[0][1])
+
+        self._sort_formats(formats)
+
+        title = self._html_search_regex(
             (r'<h1>([^<]+)</h1>', r'<title>([^<]+) - VidLii<'), webpage,
             'title')
 
@@ -82,9 +125,9 @@ class VidLiiIE(InfoExtractor):
             default=None) or self._search_regex(
             r'duration\s*:\s*(\d+)', webpage, 'duration', fatal=False))
 
-        view_count = int_or_none(self._search_regex(
-            (r'<strong>(\d+)</strong> views',
-             r'Views\s*:\s*<strong>(\d+)</strong>'),
+        view_count = str_to_int(self._html_search_regex(
+            (r'<strong>([\d,.]+)</strong> views',
+             r'Views\s*:\s*<strong>([\d,.]+)</strong>'),
             webpage, 'view count', fatal=False))
 
         comment_count = int_or_none(self._search_regex(
@@ -109,7 +152,7 @@ class VidLiiIE(InfoExtractor):
 
         return {
             'id': video_id,
-            'url': video_url,
+            'formats': formats,
             'title': title,
             'description': description,
             'thumbnail': thumbnail,