Merge 637ccf3523 into c699bafc50

Closes #11537

Authored by: bashonly

test/test_traversal.py

@@ -481,7 +481,7 @@ class TestTraversalHelpers:
             'id': 'name',
             'data': 'content',
             'url': 'url',
-        }, all, {subs_list_to_dict}]) == {
+        }, all, {subs_list_to_dict(lang=None)}]) == {
             'de': [{'url': 'https://example.com/subs/de.ass'}],
             'en': [{'data': 'content'}],
         }, 'subs with mandatory items missing should be filtered'
@@ -507,6 +507,54 @@ class TestTraversalHelpers:
             {'url': 'https://example.com/subs/en1', 'ext': 'ext'},
             {'url': 'https://example.com/subs/en2', 'ext': 'ext'},
         ]}, '`quality` key should sort subtitle list accordingly'
+        assert traverse_obj([
+            {'name': 'de', 'url': 'https://example.com/subs/de.ass'},
+            {'name': 'de'},
+            {'name': 'en', 'content': 'content'},
+            {'url': 'https://example.com/subs/en'},
+        ], [..., {
+            'id': 'name',
+            'url': 'url',
+            'data': 'content',
+        }, all, {subs_list_to_dict(lang='en')}]) == {
+            'de': [{'url': 'https://example.com/subs/de.ass'}],
+            'en': [
+                {'data': 'content'},
+                {'url': 'https://example.com/subs/en'},
+            ],
+        }, 'optionally provided lang should be used if no id available'
+        assert traverse_obj([
+            {'name': 1, 'url': 'https://example.com/subs/de1'},
+            {'name': {}, 'url': 'https://example.com/subs/de2'},
+            {'name': 'de', 'ext': 1, 'url': 'https://example.com/subs/de3'},
+            {'name': 'de', 'ext': {}, 'url': 'https://example.com/subs/de4'},
+        ], [..., {
+            'id': 'name',
+            'url': 'url',
+            'ext': 'ext',
+        }, all, {subs_list_to_dict(lang=None)}]) == {
+            'de': [
+                {'url': 'https://example.com/subs/de3'},
+                {'url': 'https://example.com/subs/de4'},
+            ],
+        }, 'non str types should be ignored for id and ext'
+        assert traverse_obj([
+            {'name': 1, 'url': 'https://example.com/subs/de1'},
+            {'name': {}, 'url': 'https://example.com/subs/de2'},
+            {'name': 'de', 'ext': 1, 'url': 'https://example.com/subs/de3'},
+            {'name': 'de', 'ext': {}, 'url': 'https://example.com/subs/de4'},
+        ], [..., {
+            'id': 'name',
+            'url': 'url',
+            'ext': 'ext',
+        }, all, {subs_list_to_dict(lang='de')}]) == {
+            'de': [
+                {'url': 'https://example.com/subs/de1'},
+                {'url': 'https://example.com/subs/de2'},
+                {'url': 'https://example.com/subs/de3'},
+                {'url': 'https://example.com/subs/de4'},
+            ],
+        }, 'non str types should be replaced by default id'
 
     def test_trim_str(self):
         with pytest.raises(TypeError):
@@ -525,7 +573,7 @@ class TestTraversalHelpers:
     def test_unpack(self):
         assert unpack(lambda *x: ''.join(map(str, x)))([1, 2, 3]) == '123'
         assert unpack(join_nonempty)([1, 2, 3]) == '1-2-3'
-        assert unpack(join_nonempty(delim=' '))([1, 2, 3]) == '1 2 3'
+        assert unpack(join_nonempty, delim=' ')([1, 2, 3]) == '1 2 3'
         with pytest.raises(TypeError):
             unpack(join_nonempty)()
         with pytest.raises(TypeError):

test/test_utils.py

@@ -72,7 +72,6 @@ from yt_dlp.utils import (
     intlist_to_bytes,
     iri_to_uri,
     is_html,
-    join_nonempty,
     js_to_json,
     limit_length,
     locked_file,
@@ -2158,10 +2157,6 @@ Line 1
         assert int_or_none(v=10) == 10, 'keyword passed positional should call function'
         assert int_or_none(scale=0.1)(10) == 100, 'call after partial application should call the function'
 
-        assert callable(join_nonempty(delim=', ')), 'varargs positional should apply partially'
-        assert callable(join_nonempty()), 'varargs positional should apply partially'
-        assert join_nonempty(None, delim=', ') == '', 'passed varargs should call the function'
-
 
 if __name__ == '__main__':
     unittest.main()

yt_dlp/YoutubeDL.py

@@ -4381,7 +4381,9 @@ class YoutubeDL:
             return None
 
         for idx, t in list(enumerate(thumbnails))[::-1]:
-            thumb_ext = (f'{t["id"]}.' if multiple else '') + determine_ext(t['url'], 'jpg')
+            thumb_ext = t.get('ext') or determine_ext(t['url'], 'jpg')
+            if multiple:
+                thumb_ext = f'{t["id"]}.{thumb_ext}'
             thumb_display_id = f'{label} thumbnail {t["id"]}'
             thumb_filename = replace_extension(filename, thumb_ext, info_dict.get('ext'))
             thumb_filename_final = replace_extension(thumb_filename_base, thumb_ext, info_dict.get('ext'))

yt_dlp/extractor/afreecatv.py

@@ -66,6 +66,14 @@ class AfreecaTVBaseIE(InfoExtractor):
             extensions={'legacy_ssl': True}), display_id,
             'Downloading API JSON', 'Unable to download API JSON')
 
+    @staticmethod
+    def _fixup_thumb(thumb_url):
+        if not url_or_none(thumb_url):
+            return None
+        # Core would determine_ext as 'php' from the url, so we need to provide the real ext
+        # See: https://github.com/yt-dlp/yt-dlp/issues/11537
+        return [{'url': thumb_url, 'ext': 'jpg'}]
+
 
 class AfreecaTVIE(AfreecaTVBaseIE):
     IE_NAME = 'soop'
@@ -155,7 +163,7 @@ class AfreecaTVIE(AfreecaTVBaseIE):
             'uploader': ('writer_nick', {str}),
             'uploader_id': ('bj_id', {str}),
             'duration': ('total_file_duration', {int_or_none(scale=1000)}),
-            'thumbnail': ('thumb', {url_or_none}),
+            'thumbnails': ('thumb', {self._fixup_thumb}),
         })
 
         entries = []
@@ -226,8 +234,7 @@ class AfreecaTVCatchStoryIE(AfreecaTVBaseIE):
 
         return self.playlist_result(self._entries(data), video_id)
 
-    @staticmethod
-    def _entries(data):
+    def _entries(self, data):
         # 'files' is always a list with 1 element
         yield from traverse_obj(data, (
             'data', lambda _, v: v['story_type'] == 'catch',
@@ -238,7 +245,7 @@ class AfreecaTVCatchStoryIE(AfreecaTVBaseIE):
                 'title': ('title', {str}),
                 'uploader': ('writer_nick', {str}),
                 'uploader_id': ('writer_id', {str}),
-                'thumbnail': ('thumb', {url_or_none}),
+                'thumbnails': ('thumb', {self._fixup_thumb}),
                 'timestamp': ('write_timestamp', {int_or_none}),
             }))
 

yt_dlp/extractor/common.py

@@ -279,6 +279,7 @@ class InfoExtractor:
     thumbnails:     A list of dictionaries, with the following entries:
                         * "id" (optional, string) - Thumbnail format ID
                         * "url"
+                        * "ext" (optional, string) - actual image extension if not given in URL
                         * "preference" (optional, int) - quality of the image
                         * "width" (optional, int)
                         * "height" (optional, int)

yt_dlp/extractor/manyvids.py

@@ -1,113 +1,100 @@
-import re
+
 
 from .common import InfoExtractor
-from ..utils import (
-    determine_ext,
-    extract_attributes,
-    int_or_none,
-    str_to_int,
-    url_or_none,
-    urlencode_postdata,
-)
+from .. import traverse_obj
+from ..utils import determine_ext, int_or_none, parse_count, parse_duration, parse_iso8601, url_or_none
 
 
 class ManyVidsIE(InfoExtractor):
-    _WORKING = False
+    _WORKING = True
     _VALID_URL = r'(?i)https?://(?:www\.)?manyvids\.com/video/(?P<id>\d+)'
-    _TESTS = [{
-        # preview video
-        'url': 'https://www.manyvids.com/Video/133957/everthing-about-me/',
-        'md5': '03f11bb21c52dd12a05be21a5c7dcc97',
-        'info_dict': {
-            'id': '133957',
-            'ext': 'mp4',
-            'title': 'everthing about me (Preview)',
-            'uploader': 'ellyxxix',
-            'view_count': int,
-            'like_count': int,
+    _TESTS = [
+        {
+            # Dead preview video
+            'skip': True,
+            'url': 'https://www.manyvids.com/Video/133957/everthing-about-me/',
+            'md5': '03f11bb21c52dd12a05be21a5c7dcc97',
+            'info_dict': {
+                'id': '133957',
+                'ext': 'mp4',
+                'title': 'everthing about me (Preview)',
+                'uploader': 'ellyxxix',
+                'view_count': int,
+                'like_count': int,
+            },
         },
-    }, {
-        # full video
-        'url': 'https://www.manyvids.com/Video/935718/MY-FACE-REVEAL/',
-        'md5': 'bb47bab0e0802c2a60c24ef079dfe60f',
-        'info_dict': {
-            'id': '935718',
-            'ext': 'mp4',
-            'title': 'MY FACE REVEAL',
-            'description': 'md5:ec5901d41808b3746fed90face161612',
-            'uploader': 'Sarah Calanthe',
-            'view_count': int,
-            'like_count': int,
+        {
+            # preview video
+            'url': 'https://www.manyvids.com/Video/530341/mv-tips-tricks',
+            'md5': '738dc723f7735ee9602f7ea352a6d058',
+            'info_dict': {
+                'id': '530341',
+                'ext': 'mp4',
+                'title': 'MV Tips &amp; Tricks (Preview)',
+                'description': 'md5:c3bae98c0f9453237c28b0f8795d9f83',
+                'thumbnail': 'https://cdn5.manyvids.com/php_uploads/video_images/DestinyDiaz/thumbs/thumb_Hs26ATOO7fcZaI9sx3XT_screenshot_001.jpg',
+                'uploader': 'DestinyDiaz',
+                'view_count': int,
+                'like_count': int,
+                'release_timestamp': 1508419904,
+                'tags': ['AdultSchool', 'BBW', 'SFW', 'TeacherFetish'],
+                'release_date': '20171019',
+                'duration': 3167.0,
+            },
         },
-    }]
+        {
+            # full video
+            'url': 'https://www.manyvids.com/Video/935718/MY-FACE-REVEAL/',
+            'md5': 'bb47bab0e0802c2a60c24ef079dfe60f',
+            'info_dict': {
+                'id': '935718',
+                'ext': 'mp4',
+                'title': 'MY FACE REVEAL',
+                'description': 'md5:ec5901d41808b3746fed90face161612',
+                'thumbnail': 'https://ods.manyvids.com/1001061960/3aa5397f2a723ec4597e344df66ab845/screenshots/thumbs/custom_1_180_5be09c1dcce03.jpg',
+                'uploader': 'Sarah Calanthe',
+                'view_count': int,
+                'like_count': int,
+                'release_date': '20181110',
+                'tags': ['EyeContact', 'Interviews', 'MaskFetish', 'MouthFetish', 'Redhead'],
+                'release_timestamp': 1541851200,
+                'duration': 224.0,
+            },
+        },
+    ]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
 
-        real_url = f'https://www.manyvids.com/video/{video_id}/gtm.js'
-        try:
-            webpage = self._download_webpage(real_url, video_id)
-        except Exception:
-            # probably useless fallback
-            webpage = self._download_webpage(url, video_id)
+        info = traverse_obj(
+            self._download_json(f'https://www.manyvids.com/bff/store/video/{video_id}', video_id),
+            ('data', {dict})) or {}
 
-        info = self._search_regex(
-            r'''(<div\b[^>]*\bid\s*=\s*(['"])pageMetaDetails\2[^>]*>)''',
-            webpage, 'meta details', default='')
-        info = extract_attributes(info)
-
-        player = self._search_regex(
-            r'''(<div\b[^>]*\bid\s*=\s*(['"])rmpPlayerStream\2[^>]*>)''',
-            webpage, 'player details', default='')
-        player = extract_attributes(player)
+        video_urls = traverse_obj(
+            self._download_json(f'https://www.manyvids.com/bff/store/video/{video_id}/private', video_id),
+            ('data', {dict})) or {}
 
         video_urls_and_ids = (
-            (info.get('data-meta-video'), 'video'),
-            (player.get('data-video-transcoded'), 'transcoded'),
-            (player.get('data-video-filepath'), 'filepath'),
-            (self._og_search_video_url(webpage, secure=False, default=None), 'og_video'),
+            (traverse_obj(video_urls, ('teaser', 'filepath')), 'preview'),
+            (video_urls.get('transcodedFilepath'), 'transcoded'),
+            (video_urls.get('filepath'), 'filepath'),
         )
 
-        def txt_or_none(s, default=None):
-            return (s.strip() or default) if isinstance(s, str) else default
+        title = info.get('title')
+        uploader = traverse_obj(info, ('model', 'displayName'))
+        description = info.get('description')
+        likes = parse_count(info.get('likes'))
+        views = parse_count(info.get('views'))
+        thumbnail = url_or_none(info.get('screenshot')) or url_or_none(info.get('thumbnail'))
+        release_timestamp = parse_iso8601(info.get('launchDate'))
+        duration = parse_duration(info.get('videoDuration'))
+        tags = [t.get('label') for t in info.get('tagList')]
 
-        uploader = txt_or_none(info.get('data-meta-author'))
-
-        def mung_title(s):
-            if uploader:
-                s = re.sub(rf'^\s*{re.escape(uploader)}\s+[|-]', '', s)
-            return txt_or_none(s)
-
-        title = (
-            mung_title(info.get('data-meta-title'))
-            or self._html_search_regex(
-                (r'<span[^>]+class=["\']item-title[^>]+>([^<]+)',
-                 r'<h2[^>]+class=["\']h2 m-0["\'][^>]*>([^<]+)'),
-                webpage, 'title', default=None)
-            or self._html_search_meta(
-                'twitter:title', webpage, 'title', fatal=True))
-
-        title = re.sub(r'\s*[|-]\s+ManyVids\s*$', '', title) or title
-
-        if any(p in webpage for p in ('preview_videos', '_preview.mp4')):
+        # If the video formats JSON only contains a teaser object, then it is a preview
+        if video_urls.get('teaser') and not video_urls.get('filepath'):
             title += ' (Preview)'
-
-        mv_token = self._search_regex(
-            r'data-mvtoken=(["\'])(?P<value>(?:(?!\1).)+)\1', webpage,
-            'mv token', default=None, group='value')
-
-        if mv_token:
-            # Sets some cookies
-            self._download_webpage(
-                'https://www.manyvids.com/includes/ajax_repository/you_had_me_at_hello.php',
-                video_id, note='Setting format cookies', fatal=False,
-                data=urlencode_postdata({
-                    'mvtoken': mv_token,
-                    'vid': video_id,
-                }), headers={
-                    'Referer': url,
-                    'X-Requested-With': 'XMLHttpRequest',
-                })
+            self.report_warning(
+                f'Only extracting preview. Video may be paid or subscription only. {self._login_hint()}')
 
         formats = []
         for v_url, fmt in video_urls_and_ids:
@@ -130,33 +117,21 @@ class ManyVidsIE(InfoExtractor):
             if f.get('height') is None:
                 f['height'] = int_or_none(
                     self._search_regex(r'_(\d{2,3}[02468])_', f['url'], 'video height', default=None))
-            if '/preview/' in f['url']:
-                f['format_id'] = '_'.join(filter(None, (f.get('format_id'), 'preview')))
+            if 'preview' in f['format_id']:
                 f['preference'] = -10
             if 'transcoded' in f['format_id']:
                 f['preference'] = f.get('preference', -1) - 1
 
-        def get_likes():
-            likes = self._search_regex(
-                rf'''(<a\b[^>]*\bdata-id\s*=\s*(['"]){video_id}\2[^>]*>)''',
-                webpage, 'likes', default='')
-            likes = extract_attributes(likes)
-            return int_or_none(likes.get('data-likes'))
-
-        def get_views():
-            return str_to_int(self._html_search_regex(
-                r'''(?s)<span\b[^>]*\bclass\s*=["']views-wrapper\b[^>]+>.+?<span\b[^>]+>\s*(\d[\d,.]*)\s*</span>''',
-                webpage, 'view count', default=None))
-
         return {
             'id': video_id,
             'title': title,
             'formats': formats,
-            'description': txt_or_none(info.get('data-meta-description')),
-            'uploader': txt_or_none(info.get('data-meta-author')),
-            'thumbnail': (
-                url_or_none(info.get('data-meta-image'))
-                or url_or_none(player.get('data-video-screenshot'))),
-            'view_count': get_views(),
-            'like_count': get_likes(),
+            'description': description,
+            'uploader': uploader,
+            'thumbnail': thumbnail,
+            'view_count': views,
+            'like_count': likes,
+            'release_timestamp': release_timestamp,
+            'duration': duration,
+            'tags': tags,
         }

yt_dlp/utils/_utils.py

@@ -216,7 +216,7 @@ def partial_application(func):
     sig = inspect.signature(func)
     required_args = [
         param.name for param in sig.parameters.values()
-        if param.kind in (inspect.Parameter.POSITIONAL_ONLY, inspect.Parameter.POSITIONAL_OR_KEYWORD, inspect.Parameter.VAR_POSITIONAL)
+        if param.kind in (inspect.Parameter.POSITIONAL_ONLY, inspect.Parameter.POSITIONAL_OR_KEYWORD)
         if param.default is inspect.Parameter.empty
     ]
 
@@ -4837,7 +4837,6 @@ def number_of_digits(number):
     return len('%d' % number)
 
 
-@partial_application
 def join_nonempty(*values, delim='-', from_dict=None):
     if from_dict is not None:
         values = (traversal.traverse_obj(from_dict, variadic(v)) for v in values)

yt_dlp/utils/traversal.py

@@ -332,14 +332,14 @@ class _RequiredError(ExtractorError):
 
 
 @typing.overload
-def subs_list_to_dict(*, ext: str | None = None) -> collections.abc.Callable[[list[dict]], dict[str, list[dict]]]: ...
+def subs_list_to_dict(*, lang: str | None = 'und', ext: str | None = None) -> collections.abc.Callable[[list[dict]], dict[str, list[dict]]]: ...
 
 
 @typing.overload
-def subs_list_to_dict(subs: list[dict] | None, /, *, ext: str | None = None) -> dict[str, list[dict]]: ...
+def subs_list_to_dict(subs: list[dict] | None, /, *, lang: str | None = 'und', ext: str | None = None) -> dict[str, list[dict]]: ...
 
 
-def subs_list_to_dict(subs: list[dict] | None = None, /, *, ext=None):
+def subs_list_to_dict(subs: list[dict] | None = None, /, *, lang='und', ext=None):
     """
     Convert subtitles from a traversal into a subtitle dict.
     The path should have an `all` immediately before this function.
@@ -352,7 +352,7 @@ def subs_list_to_dict(subs: list[dict] | None = None, /, *, ext=None):
     `quality`  The sort order for each subtitle
     """
     if subs is None:
-        return functools.partial(subs_list_to_dict, ext=ext)
+        return functools.partial(subs_list_to_dict, lang=lang, ext=ext)
 
     result = collections.defaultdict(list)
 
@@ -360,10 +360,16 @@ def subs_list_to_dict(subs: list[dict] | None = None, /, *, ext=None):
         if not url_or_none(sub.get('url')) and not sub.get('data'):
             continue
         sub_id = sub.pop('id', None)
-        if sub_id is None:
-            continue
-        if ext is not None and not sub.get('ext'):
-            sub['ext'] = ext
+        if not isinstance(sub_id, str):
+            if not lang:
+                continue
+            sub_id = lang
+        sub_ext = sub.get('ext')
+        if not isinstance(sub_ext, str):
+            if not ext:
+                sub.pop('ext', None)
+            else:
+                sub['ext'] = ext
         result[sub_id].append(sub)
     result = dict(result)
 
@@ -452,9 +458,9 @@ def trim_str(*, start=None, end=None):
     return trim
 
 
-def unpack(func):
+def unpack(func, **kwargs):
     @functools.wraps(func)
-    def inner(items, **kwargs):
+    def inner(items):
         return func(*items, **kwargs)
 
     return inner