Merge fe274adf41 into eb15fd5a32

Closes #3937
Authored by: krichbanana, pzhlkj6612

Co-authored-by: Mozi <29089388+pzhlkj6612@users.noreply.github.com>

yt_dlp/extractor/_extractors.py

@@ -946,6 +946,10 @@ from .kaltura import KalturaIE
 from .kankanews import KankaNewsIE
 from .karaoketv import KaraoketvIE
 from .kelbyone import KelbyOneIE
+from .kenh14 import (
+    Kenh14PlaylistIE,
+    Kenh14VideoIE,
+)
 from .khanacademy import (
     KhanAcademyIE,
     KhanAcademyUnitIE,
@@ -2617,3 +2621,4 @@ from .zingmp3 import (
 )
 from .zoom import ZoomIE
 from .zype import ZypeIE
+from .lazy import LazyExtractorIE

yt_dlp/extractor/chaturbate.py

@@ -79,7 +79,7 @@ class ChaturbateIE(InfoExtractor):
             'formats': self._extract_m3u8_formats(m3u8_url, video_id, ext='mp4', live=True),
         }
 
-    def _extract_from_webpage(self, video_id, tld):
+    def _extract_from_html(self, video_id, tld):
         webpage = self._download_webpage(
             f'https://chaturbate.{tld}/{video_id}/', video_id,
             headers=self.geo_verification_headers(), impersonate=True)
@@ -151,4 +151,4 @@ class ChaturbateIE(InfoExtractor):
 
     def _real_extract(self, url):
         video_id, tld = self._match_valid_url(url).group('id', 'tld')
-        return self._extract_from_api(video_id, tld) or self._extract_from_webpage(video_id, tld)
+        return self._extract_from_api(video_id, tld) or self._extract_from_html(video_id, tld)

yt_dlp/extractor/kenh14.py

@@ -0,0 +1,160 @@
+from .common import InfoExtractor
+from ..utils import (
+    clean_html,
+    extract_attributes,
+    get_element_by_class,
+    get_element_html_by_attribute,
+    get_elements_html_by_class,
+    int_or_none,
+    parse_duration,
+    parse_iso8601,
+    remove_start,
+    strip_or_none,
+    unescapeHTML,
+    update_url,
+    url_or_none,
+)
+from ..utils.traversal import traverse_obj
+
+
+class Kenh14VideoIE(InfoExtractor):
+    _VALID_URL = r'https?://video\.kenh14\.vn/(?:video/)?[\w-]+-(?P<id>[0-9]+)\.chn'
+    _TESTS = [{
+        'url': 'https://video.kenh14.vn/video/mo-hop-iphone-14-pro-max-nguon-unbox-therapy-316173.chn',
+        'md5': '1ed67f9c3a1e74acf15db69590cf6210',
+        'info_dict': {
+            'id': '316173',
+            'ext': 'mp4',
+            'title': 'Video mở hộp iPhone 14 Pro Max (Nguồn: Unbox Therapy)',
+            'description': 'Video mở hộp iPhone 14 Pro MaxVideo mở hộp iPhone 14 Pro Max (Nguồn: Unbox Therapy)',
+            'thumbnail': r're:^https?://videothumbs\.mediacdn\.vn/.*\.jpg$',
+            'tags': [],
+            'uploader': 'Unbox Therapy',
+            'upload_date': '20220517',
+            'view_count': int,
+            'duration': 722.86,
+            'timestamp': 1652764468,
+        },
+    }, {
+        'url': 'https://video.kenh14.vn/video-316174.chn',
+        'md5': '2b41877d2afaf4a3f487ceda8e5c7cbd',
+        'info_dict': {
+            'id': '316174',
+            'ext': 'mp4',
+            'title': 'Khoảnh khắc VĐV nằm gục khóc sau chiến thắng: 7 năm trời Việt Nam mới có HCV kiếm chém nữ, chỉ có 8 tháng để khổ luyện trước khi lên sàn đấu',
+            'description': 'md5:de86aa22e143e2b277bce8ec9c6f17dc',
+            'thumbnail': r're:^https?://videothumbs\.mediacdn\.vn/.*\.jpg$',
+            'tags': [],
+            'upload_date': '20220517',
+            'view_count': int,
+            'duration': 70.04,
+            'timestamp': 1652766021,
+        },
+    }, {
+        'url': 'https://video.kenh14.vn/0-344740.chn',
+        'md5': 'b843495d5e728142c8870c09b46df2a9',
+        'info_dict': {
+            'id': '344740',
+            'ext': 'mov',
+            'title': 'Kỳ Duyên đầy căng thẳng trong buổi ra quân đi Miss Universe, nghi thức tuyên thuệ lần đầu xuất hiện gây nhiều tranh cãi',
+            'description': 'md5:2a2dbb4a7397169fb21ee68f09160497',
+            'thumbnail': r're:^https?://kenh14cdn\.com/.*\.jpg$',
+            'tags': ['kỳ duyên', 'Kỳ Duyên tuyên thuệ', 'miss universe'],
+            'uploader': 'Quang Vũ',
+            'upload_date': '20241024',
+            'view_count': int,
+            'duration': 198.88,
+            'timestamp': 1729741590,
+        },
+    }]
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        webpage = self._download_webpage(url, video_id)
+
+        attrs = extract_attributes(get_element_html_by_attribute('type', 'VideoStream', webpage) or '')
+        direct_url = attrs['data-vid']
+
+        metadata = self._download_json(
+            'https://api.kinghub.vn/video/api/v1/detailVideoByGet?FileName={}'.format(
+                remove_start(direct_url, 'kenh14cdn.com/')), video_id, fatal=False)
+
+        formats = [{'url': f'https://{direct_url}', 'format_id': 'http', 'quality': 1}]
+        subtitles = {}
+        video_data = self._download_json(
+            f'https://{direct_url}.json', video_id, note='Downloading video data', fatal=False)
+        if hls_url := traverse_obj(video_data, ('hls', {url_or_none})):
+            fmts, subs = self._extract_m3u8_formats_and_subtitles(
+                hls_url, video_id, m3u8_id='hls', fatal=False)
+            formats.extend(fmts)
+            self._merge_subtitles(subs, target=subtitles)
+        if dash_url := traverse_obj(video_data, ('mpd', {url_or_none})):
+            fmts, subs = self._extract_mpd_formats_and_subtitles(
+                dash_url, video_id, mpd_id='dash', fatal=False)
+            formats.extend(fmts)
+            self._merge_subtitles(subs, target=subtitles)
+
+        return {
+            **traverse_obj(metadata, {
+                'duration': ('duration', {parse_duration}),
+                'uploader': ('author', {strip_or_none}),
+                'timestamp': ('uploadtime', {parse_iso8601(delimiter=' ')}),
+                'view_count': ('views', {int_or_none}),
+            }),
+            'id': video_id,
+            'title': (
+                traverse_obj(metadata, ('title', {strip_or_none}))
+                or clean_html(self._og_search_title(webpage))
+                or clean_html(get_element_by_class('vdbw-title', webpage))),
+            'formats': formats,
+            'subtitles': subtitles,
+            'description': (
+                clean_html(self._og_search_description(webpage))
+                or clean_html(get_element_by_class('vdbw-sapo', webpage))),
+            'thumbnail': (self._og_search_thumbnail(webpage) or attrs.get('data-thumb')),
+            'tags': traverse_obj(self._html_search_meta('keywords', webpage), (
+                {lambda x: x.split(';')}, ..., filter)),
+        }
+
+
+class Kenh14PlaylistIE(InfoExtractor):
+    _VALID_URL = r'https?://video\.kenh14\.vn/playlist/[\w-]+-(?P<id>[0-9]+)\.chn'
+    _TESTS = [{
+        'url': 'https://video.kenh14.vn/playlist/tran-tinh-naked-love-mua-2-71.chn',
+        'info_dict': {
+            'id': '71',
+            'title': 'Trần Tình (Naked love) mùa 2',
+            'description': 'md5:e9522339304956dea931722dd72eddb2',
+            'thumbnail': r're:^https?://kenh14cdn\.com/.*\.png$',
+        },
+        'playlist_count': 9,
+    }, {
+        'url': 'https://video.kenh14.vn/playlist/0-72.chn',
+        'info_dict': {
+            'id': '72',
+            'title': 'Lau Lại Đầu Từ',
+            'description': 'Cùng xem xưa và nay có gì khác biệt nhé!',
+            'thumbnail': r're:^https?://kenh14cdn\.com/.*\.png$',
+        },
+        'playlist_count': 6,
+    }]
+
+    def _real_extract(self, url):
+        playlist_id = self._match_id(url)
+        webpage = self._download_webpage(url, playlist_id)
+
+        category_detail = get_element_by_class('category-detail', webpage) or ''
+        embed_info = traverse_obj(
+            self._yield_json_ld(webpage, playlist_id),
+            (lambda _, v: v['name'] and v['alternateName'], any)) or {}
+
+        return self.playlist_from_matches(
+            get_elements_html_by_class('video-item', webpage), playlist_id,
+            (clean_html(get_element_by_class('name', category_detail)) or unescapeHTML(embed_info.get('name'))),
+            getter=lambda x: 'https://video.kenh14.vn/video/video-{}.chn'.format(extract_attributes(x)['data-id']),
+            ie=Kenh14VideoIE, playlist_description=(
+                clean_html(get_element_by_class('description', category_detail))
+                or unescapeHTML(embed_info.get('alternateName'))),
+            thumbnail=traverse_obj(
+                self._og_search_thumbnail(webpage),
+                ({url_or_none}, {update_url(query=None)})))

yt_dlp/extractor/lazy.py

@@ -0,0 +1,33 @@
+from .common import InfoExtractor
+from ..utils.lazy import lazy_ie, lazy_fields
+
+
+@lazy_ie
+class LazyExtractorIE(InfoExtractor):
+    IE_NAME = 'lazy'
+    _VALID_URL = r"lazy://(?P<id>.*)"
+
+    def _lazy_webpage(self, storage):
+        return self._download_webpage(storage.url, storage.id)
+
+    @lazy_fields("creator")
+    def _extract_other(self, storage):
+        self.to_screen("Extracting something else from webpage")
+        return {
+            "creator": storage.webpage.partition(" - ")[0],
+        }
+
+    @lazy_fields("title", "description")
+    def _extract_website(self, storage):
+        self.to_screen("Extracting title and description from webpage")
+        title, _, description = storage.webpage.partition("\n")
+
+        return {
+            "title": title,
+            "description": description,
+        }
+
+    # Fake downloading the webpage for testing purposes
+    def _download_webpage(self, url_or_request, video_id, *args, **kwargs):
+        self.to_screen(f"[{video_id}] Downloaded webpage ({url_or_request})")
+        return "<creator> - Fake Webpage title\nThis is the description.\n..."

yt_dlp/utils/lazy.py

@@ -0,0 +1,153 @@
+from __future__ import annotations
+
+import functools
+from collections.abc import MutableMapping
+
+from ..utils import try_call
+from ..extractor.common import InfoExtractor
+
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from typing import Callable, Any
+
+
+class _LazyStorage:
+    def __init__(self, ie, **kwargs):
+        self._ie = ie
+        self._cache = kwargs
+
+    def __setattr__(self, name, value, /) -> None:
+        if name.startswith("_"):
+            super().__setattr__(name, value)
+        else:
+            self._cache[name] = value
+
+    def __getattr__(self, name: str):
+        if name in self._cache:
+            return self._cache[name]
+
+        resolver = getattr(self._ie, f"_lazy_{name}")
+        result = try_call(resolver, args=(self,))
+        self._cache[name] = result
+        return result
+
+    def __delattr__(self, name: str) -> None:
+        if name.startswith("_"):
+            super().__delattr__(name)
+        elif name in self._cache:
+            del self._cache[name]
+
+
+class _LazyInfoDict(MutableMapping):
+    def __init__(self, data: dict, lazy: dict, ie: InfoExtractor, **kwargs):
+        self._data = data
+        self._lazy = lazy
+        self._ie = ie
+        self._storage = _LazyStorage(self._ie, **kwargs)
+
+        for key in self._data.keys() & self._lazy.keys():
+            del self._lazy[key]
+
+        self._data.update(dict.fromkeys(self._lazy.keys()))
+
+    def __contains__(self, key):
+        return key in self._data
+
+    def __getitem__(self, key):
+        if key in self._lazy:
+            compute_func = self._lazy[key]
+
+            # updates = try_call(compute_func, args=(self._storage,), expected_type=dict) or {}
+            updates = compute_func(self._ie, self._storage)
+            self._data.update(updates)
+            for field in updates:
+                self._lazy.pop(field, None)
+
+            fields = getattr(compute_func, lazy_fields._field_name, None) or ()
+            for field in fields:
+                self._lazy.pop(field, None)
+
+        return self._data[key]
+
+    def __setitem__(self, key, value):
+        if key in self._lazy:
+            del self._lazy[key]
+
+        self._data[key] = value
+
+    def __delitem__(self, key):
+        if key in self._lazy:
+            del self._lazy[key]
+
+        del self._data[key]
+
+    def __iter__(self):
+        return iter(self._data)
+
+    def __len__(self):
+        return len(self._data)
+
+    def __repr__(self):
+        if self._lazy:
+            lazy = ", ".join(f"{key!r}: ..." for key in self._lazy.keys())
+            data = ", ".join(f"{key!r}: {value!r}" for key, value in self._data.items() if key not in self._lazy)
+            data = f"{{{data}}}, lazy={{{lazy}}}"
+        else:
+            data = f"{self._data!r}"
+        return f"{type(self).__name__}({data})"
+
+
+def _default_lazy_extract(self, url):
+    return dict(id=self._match_id(url))
+
+
+def lazy_ie(klass: type[InfoExtractor] | None = None, /):
+    if not klass:
+        return lazy_ie
+
+    _old_extract = klass._real_extract
+    if _old_extract is InfoExtractor._real_extract:
+        _old_extract = _default_lazy_extract
+
+    lazy_members = {}
+    for name in dir(klass):
+        if not name.startswith("_"):
+            continue
+        func = getattr(klass, name)
+        fields = getattr(func, lazy_fields._field_name, None)
+        if not isinstance(fields, tuple):
+            continue
+
+        for field in fields:
+            lazy_members[field] = func
+
+    @functools.wraps(klass._real_extract)
+    def _real_extract(self, url):
+        result = _old_extract(self, url)
+        assert isinstance(result, dict), 'Lazy extractors need to return a dict'
+        return _LazyInfoDict(result, lazy_members, self, url=url, **result)
+
+    klass._real_extract = _real_extract
+    return klass
+
+
+def lazy_fields(*fields: str) -> Callable[[Callable[[Any, _LazyStorage], dict[str, Any]]], Callable[[Any, _LazyStorage], dict[str, Any]]]:
+    def _lazy_fields(func):
+        setattr(func, lazy_fields._field_name, fields)
+        return func
+
+    return _lazy_fields
+
+
+lazy_fields._field_name = "_lazy_fields"
+
+if __name__ == '__main__':
+    from yt_dlp import YoutubeDL
+
+    with YoutubeDL() as ydl:
+        result = ydl.extract_info("lazy://<URL>", process=False)
+        assert result
+
+        for name in "id", "title", "creator", "description":
+            print(f"{name:<10} = {result[name]!r}")