split extractors

yt_dlp/extractor/_extractors.py

@@ -1906,7 +1906,12 @@ from .syvdk import SYVDKIE
 from .syfy import SyfyIE
 from .sztvhu import SztvHuIE
 from .tagesschau import TagesschauIE
-from .taptap import TapTapIE
+from .taptap import (
+    TapTapMomentIE,
+    TapTapAppIE,
+    TapTapAppIntlIE,
+    TapTapPostIntlIE,
+)
 from .tass import TassIE
 from .tbs import TBSIE
 from .tbsjp import (

yt_dlp/extractor/taptap.py

@@ -1,20 +1,84 @@
+import random
+import re
+
 from .common import InfoExtractor
 from ..utils import (
     clean_html,
     int_or_none,
-    parse_qs,
+    join_nonempty,
-    str_or_none,
     traverse_obj,
     url_or_none,
 )
 
 
-class TapTapIE(InfoExtractor):
+class TapTapBaseIE(InfoExtractor):
-    _VALID_URL = r'https?://www\.taptap\.cn/(?P<section>moment|app)/(?P<id>\d+)'
+    _X_UA = 'V=1&PN=WebApp&LANG=zh_CN&VN_CODE=102&LOC=CN&PLT=PC&DS=Android&UID={uuid}&OS=Windows&OSV=10&DT=PC'
+    _VIDEO_API = 'https://www.taptap.cn/webapiv2/video-resource/v1/multi-get'
+    _INFO_API = None
+    _INFO_QUERY_KEY = 'id'
+    _DATA_PATH = ('data')
+    _ID_PATH = None
+    _META_PATH = None
+
+    def _get_api(self, url, video_id, query, **kwargs):
+        rand_hex = lambda digits: ''.join(f'{random.randint(0, 15):x}' for _ in range(digits))
+        uuid = '-'.join(rand_hex(digits) for digits in [8, 4, 4, 4, 12])
+        query = {**query, 'X-UA': self._X_UA.format(uuid=uuid)}
+        return self._download_json(url, video_id, query=query, **kwargs)
+
+    def _extract_video(self, video_id, is_intl=False):
+        video_data = self._get_api(self._VIDEO_API, video_id, query={'video_ids': video_id})['data']['list'][0]
+
+        video_url = traverse_obj(video_data, ('play_url', ('url_h265', 'url'), {url_or_none}))[0]
+        formats = self._extract_m3u8_formats(video_url, video_id)
+        for format in formats:
+            if re.search(r'^(hev|hvc|hvt)\d', format.get('vcodec', '')):
+                format['format_id'] = join_nonempty(format.get('format_id'), 'h265', delim='_')
+
+        return {
+            'id': str(video_id),
+            'formats': formats,
+            **traverse_obj(video_data, ({
+                'duration': ('info', 'duration', {int_or_none}),
+                'thumbnail': ('thumbnail', ('original_url', 'url'), {url_or_none}),
+            }), get_all=False)
+        }
+
+    def _extract_entries(self, video_ids, metainfo, list_id):
+        entries = [{**metainfo, **self._extract_video(id)} for id in set(video_ids)]
+        return self.playlist_result(entries, **metainfo, id=list_id)
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+        query = {self._INFO_QUERY_KEY: video_id}
+
+        data = traverse_obj(
+            self._get_api(self._INFO_API, video_id, query=query), self._DATA_PATH)
+
+        video_ids = traverse_obj(data, self._ID_PATH)
+        metainfo = traverse_obj(data, self._META_PATH)
+        return self._extract_entries(video_ids, metainfo, video_id)
+
+
+class TapTapIntlBase(TapTapBaseIE):
+    _X_UA = 'V=1&PN=WebAppIntl2&LANG=zh_TW&VN_CODE=115&VN=0.1.0&LOC=CN&PLT=PC&DS=Android&UID={uuid}&CURR=&DT=PC&OS=Windows&OSV=NT%208.0.0'
+    _VIDEO_API = 'https://www.taptap.io/webapiv2/video-resource/v1/multi-get'
+
+
+class TapTapMomentIE(TapTapBaseIE):
+    _VALID_URL = r'https?://www\.taptap\.cn/moment/(?P<id>\d+)'
+    _INFO_API = 'https://www.taptap.cn/webapiv2/moment/v3/detail'
+    _ID_PATH = ('moment', 'topic', (('videos', ...), 'pin_video'), 'video_id')
+    _META_PATH = ('moment', {
+        'timestamp': ('created_time', {int_or_none}),
+        'uploader': ('author', 'user', 'name', {str}),
+        'title': ('topic', 'title', {str}),
+        'description': ('topic', 'summary', {str}),
+    })
     _TESTS = [{
         'url': 'https://www.taptap.cn/moment/194618230982052443',
         'info_dict': {
-            'id': 'moment_194618230982052443',
+            'id': '194618230982052443',
             'title': '《崩坏3》开放世界「后崩坏书」新篇章 于淹没之地仰视辰星',
             'description': 'md5:cf66f7819d413641b8b28c8543f4ecda',
             'timestamp': 1633453402,
@@ -38,7 +102,7 @@ class TapTapIE(InfoExtractor):
     }, {
         'url': 'https://www.taptap.cn/moment/521630629209573493',
         'info_dict': {
-            'id': 'moment_521630629209573493',
+            'id': '521630629209573493',
             'title': '《崩坏：星穹铁道》黄泉角色PV——「你的颜色」',
             'description': 'md5:2c81245da864428c904d53ae4ad2182b',
             'timestamp': 1711425600,
@@ -59,10 +123,21 @@ class TapTapIE(InfoExtractor):
                 'thumbnail': r're:^https?://.*\.(png|jpg)',
             }
         }]
-    }, {
+    }]
+
+
+class TapTapAppIE(TapTapBaseIE):
+    _VALID_URL = r'https?://www\.taptap\.cn/app/(?P<id>\d+)'
+    _INFO_API = 'https://www.taptap.cn/webapiv2/app/v4/detail'
+    _ID_PATH = (('app_videos', 'videos'), ..., 'video_id')
+    _META_PATH = {
+        'title': ('title', {str}),
+        'description': ('description', 'text', {str}, {clean_html}),
+    }
+    _TESTS = [{
         'url': 'https://www.taptap.cn/app/168332',
         'info_dict': {
-            'id': 'app_168332',
+            'id': '168332',
             'title': '原神',
             'description': 'md5:e345f39a5fea5de2a46923f70d5f76ab',
         },
@@ -86,63 +161,49 @@ class TapTapIE(InfoExtractor):
                 'thumbnail': r're:^https?://.*\.(png|jpg)',
             }
         }]
-
     }]
 
-    def _deserialize_nuxt_data(self, serialized_nuxt):
-        for row in serialized_nuxt:
-            if isinstance(row, dict):
-                for key, value_or_ref in row.items():
-                    if isinstance(value_or_ref, int):
-                        row[key] = serialized_nuxt[value_or_ref]
-            elif isinstance(row, list):
-                for index, value_or_ref in tuple(enumerate(row)):
-                    if isinstance(value_or_ref, int):
-                        row[index] = serialized_nuxt[value_or_ref]
-        return serialized_nuxt[0]
 
-    def _extract_video(self, video_id, x_ua):
+class TapTapAppIntlIE(TapTapAppIE, TapTapIntlBase):
-        data = self._download_json(
+    _VALID_URL = r'https?://www\.taptap\.io/app/(?P<id>\d+)'
-            'https://www.taptap.cn/webapiv2/video-resource/v1/multi-get', video_id,
+    _INFO_API = 'https://www.taptap.io/webapiv2/i/app/v5/detail'
-            query={'video_ids': video_id, 'X-UA': x_ua})
+    _DATA_PATH = ('data', 'app')
 
-        video = traverse_obj(data, ('data', 'list', 0, {
-            'id': ('video_id', {str_or_none}),
-            'url': ('play_url', ('url', 'url_h265'), {url_or_none}),
-            'duration': ('info', 'duration', {int_or_none}),
-            'thumbnail': ('thumbnail', ('original_url', 'url'), {url_or_none}),
-        }), get_all=False)
-        if '.m3u8' in video['url']:
-            video['formats'] = self._extract_m3u8_formats(video.pop('url'), video_id)
-        return video
 
-    def _real_extract(self, url):
+class TapTapPostIntlIE(TapTapAppIntlIE):
-        section, list_id = self._match_valid_url(url).groups()
+    _VALID_URL = r'https?://www\.taptap\.io/post/(?P<id>\d+)'
-        list_id = f'{section}_{list_id}'
+    _INFO_API = 'https://www.taptap.io/webapiv2/creation/post/v1/detail'
-
+    _INFO_QUERY_KEY = 'id_str'
-        webpage = self._download_webpage(url, list_id)
+    _DATA_PATH = ('data', 'post')
-        nuxt_data = self._deserialize_nuxt_data(self._search_json(
+    _ID_PATH = ((('videos', ...), 'pin_video'), 'video_id')
-            r'<script[^>]+\bid=["\']__NUXT_DATA__["\'][^>]*>', webpage,
+    _META_PATH = {
-            'nuxt data', list_id, contains_pattern=r'\[(?s:.+)\]'))[1]
+        'timestamp': ('published_time', {int_or_none}),
-        x_ua = traverse_obj(nuxt_data, (
+        'uploader': ('user', 'name', {str}),
-            'state', '$sbff', ..., {lambda x: parse_qs(x)['X-UA']}, ...), get_all=False)
+        'title': ('title', {str}),
-
+        'description': ('list_fields', 'summary', {str}),
-        if section == 'moment':
+    }
-            moment_data = traverse_obj(nuxt_data, ('data', ..., 'moment'), get_all=False)
+    _TESTS = [{
-            video_ids = traverse_obj(moment_data, ('topic', (('videos', ...), 'pin_video'), 'video_id'))
+        'url': 'https://www.taptap.io/post/571785',
-            metainfo = traverse_obj(moment_data, {
+        'info_dict': {
-                'timestamp': ('created_time', {int_or_none}),
+            'id': '571785',
-                'uploader': ('author', 'user', 'name', {str}),
+            'title': 'Arknights x Rainbow Six Siege | Event PV',
-                'title': ('topic', 'title', {str}),
+            'description': 'md5:f7717c13f6d3108e22db7303e6690bf7',
-                'description': ('topic', 'summary', {str}),
+            'timestamp': 1614664951,
-            })
+            'upload_date': '20210302',
-        elif section == 'app':
+            'uploader': 'TapTap Editor',
-            video_ids = traverse_obj(nuxt_data, ('data', ..., ('app_videos', 'videos'), ..., 'video_id'))
+        },
-            metainfo = traverse_obj(nuxt_data, ('data', ..., {
+        'playlist_count': 1,
-                'title': ('title', {str}),
+        'playlist': [{
-                'description': ('description', 'text', {str}, {clean_html}),
+            'info_dict': {
-            }), get_all=False)
+                'id': '2149491903',
-
+                'ext': 'mp4',
-        entries = [self._extract_video(video_id, x_ua) for video_id in set(video_ids)]
+                'title': 'Arknights x Rainbow Six Siege | Event PV',
-
+                'description': 'md5:f7717c13f6d3108e22db7303e6690bf7',
-        return self.playlist_result([{**metainfo, **e} for e in entries], **metainfo, id=list_id)
+                'duration': 122,
+                'timestamp': 1614664951,
+                'upload_date': '20210302',
+                'uploader': 'TapTap Editor',
+                'thumbnail': r're:^https?://.*\.(png|jpg)',
+            }
+        }]
+    }]