[ie/bilibili] Fix API extraction (#13730)

Closes #12887, Closes #16962
Authored by: grqz

.github/actionlint.yml

@@ -18,3 +18,7 @@ paths:
     ignore:
       # SC1090 "Can't follow non-constant source": ignore when using `source` to activate venv
       - '.+SC1090.+'
+  .github/workflows/label-handler.yml:
+    ignore:
+      # https://github.com/rhysd/actionlint/issues/657
+      - 'unexpected key "queue" for "concurrency" section.+'

.github/workflows/label-handler.yml

@@ -0,0 +1,92 @@
+name: Label Handler
+on:
+  issues:
+    types: [labeled]
+  pull_request_target:
+    types: [labeled]  # zizmor: ignore[dangerous-triggers]
+
+permissions: {}
+
+concurrency:
+  group: label
+  cancel-in-progress: false
+  queue: max
+
+env:
+  GH_TELEMETRY: "false"
+  LLM_MESSAGE: >-
+    This contribution has been determined to be in violation of yt-dlp's
+    [**NO AI / NO LLM POLICY**](https://github.com/yt-dlp/yt-dlp/blob/master/CONTRIBUTING.md#no-ai--no-llm-policy).
+    Repeat or flagrant violations of this policy will result in a permanent ban from this repository.
+  BAD_MESSAGE: >-
+    This contribution has been determined to be in violation of yt-dlp's
+    [policy against supporting sites that are primarily used for
+    piracy](https://github.com/yt-dlp/yt-dlp/blob/master/CONTRIBUTING.md#is-the-website-primarily-used-for-piracy).
+    Please consult the [yt-dlp wiki
+    FAQ](https://github.com/yt-dlp/yt-dlp/wiki/FAQ#why-is-there-a-rule-against-websites-primarily-used-for-piracy)
+    if you have questions.
+
+jobs:
+  issue-llm:
+    name: Issue (ai-policy-violation)
+    if: github.event.issue.state == 'open' && contains(github.event.issue.labels.*.name, 'ai-policy-violation')
+    permissions:
+      issues: write  # Needed to comment on, close, and lock issues
+    runs-on: ubuntu-slim
+    env:
+      GH_TOKEN: ${{ github.token }}
+      ISSUE_NUMBER: ${{ github.event.issue.number }}
+    steps:
+      - name: Comment, close and lock issue
+        run: |
+          gh issue unlock "${ISSUE_NUMBER}" --repo "${GITHUB_REPOSITORY}" || true
+          gh issue close "${ISSUE_NUMBER}" --repo "${GITHUB_REPOSITORY}" --reason "not planned" --comment "${LLM_MESSAGE}"
+          gh issue lock "${ISSUE_NUMBER}" --repo "${GITHUB_REPOSITORY}"
+
+  issue-bad:
+    name: Issue (piracy/illegal)
+    if: github.event.issue.state == 'open' && contains(github.event.issue.labels.*.name, 'piracy/illegal')
+    permissions:
+      issues: write  # Needed to comment on, close, and lock issues
+    runs-on: ubuntu-slim
+    env:
+      GH_TOKEN: ${{ github.token }}
+      ISSUE_NUMBER: ${{ github.event.issue.number }}
+    steps:
+      - name: Comment, close and lock issue
+        run: |
+          gh issue unlock "${ISSUE_NUMBER}" --repo "${GITHUB_REPOSITORY}" || true
+          gh issue close "${ISSUE_NUMBER}" --repo "${GITHUB_REPOSITORY}" --reason "not planned" --comment "${BAD_MESSAGE}"
+          gh issue lock "${ISSUE_NUMBER}" --repo "${GITHUB_REPOSITORY}"
+
+  pr-llm:
+    name: PR (ai-policy-violation)
+    if: github.event.pull_request.state == 'open' && contains(github.event.pull_request.labels.*.name, 'ai-policy-violation')
+    permissions:
+      pull-requests: write  # Needed to comment on, close, and lock PRs
+    runs-on: ubuntu-slim
+    env:
+      GH_TOKEN: ${{ github.token }}
+      PR_NUMBER: ${{ github.event.pull_request.number }}
+    steps:
+      - name: Comment, close and lock PR
+        run: |
+          gh pr unlock "${PR_NUMBER}" --repo "${GITHUB_REPOSITORY}" || true
+          gh pr close "${PR_NUMBER}" --repo "${GITHUB_REPOSITORY}" --comment "${LLM_MESSAGE}"
+          gh pr lock "${PR_NUMBER}" --repo "${GITHUB_REPOSITORY}"
+
+  pr-bad:
+    name: PR (piracy/illegal)
+    if: github.event.pull_request.state == 'open' && contains(github.event.pull_request.labels.*.name, 'piracy/illegal')
+    permissions:
+      pull-requests: write  # Needed to comment on, close, and lock PRs
+    runs-on: ubuntu-slim
+    env:
+      GH_TOKEN: ${{ github.token }}
+      PR_NUMBER: ${{ github.event.pull_request.number }}
+    steps:
+      - name: Comment, close and lock PR
+        run: |
+          gh pr unlock "${PR_NUMBER}" --repo "${GITHUB_REPOSITORY}" || true
+          gh pr close "${PR_NUMBER}" --repo "${GITHUB_REPOSITORY}" --comment "${BAD_MESSAGE}"
+          gh pr lock "${PR_NUMBER}" --repo "${GITHUB_REPOSITORY}"

README.md

@@ -403,7 +403,7 @@ Tip: Use `CTRL`+`F` (or `Command`+`F`)  to search by keywords
                                     (default)
     --live-from-start               Download livestreams from the start.
                                     Currently experimental and only supported
-                                    for YouTube, Twitch, and TVer
+                                    for YouTube, Twitch, TVer, and mellow-fan
     --no-live-from-start            Download livestreams from the current time
                                     (default)
     --wait-for-video MIN[-MAX]      Wait for scheduled streams to become
@@ -1969,6 +1969,9 @@ The following extractors use this feature:
 #### sonylivseries
 * `sort_order`: Episode sort order for series extraction - one of `asc` (ascending, oldest first) or `desc` (descending, newest first). Default is `asc`
 
+#### streaks
+* `api_key`: API key for the `X-Streaks-Api-Key` header
+
 #### tver
 * `backend`: Backend API to use for extraction - one of `streaks` (default) or `brightcove` (deprecated)
 

yt_dlp/extractor/_extractors.py

@@ -1314,6 +1314,11 @@ from .oftv import (
 )
 from .oktoberfesttv import OktoberfestTVIE
 from .olympics import OlympicsReplayIE
+from .omnyfm import (
+    OmnyfmIE,
+    OmnyfmPlaylistIE,
+    OmnyfmShowIE,
+)
 from .on24 import On24IE
 from .ondemandkorea import (
     OnDemandKoreaIE,
@@ -1335,8 +1340,11 @@ from .opencast import (
 )
 from .openrec import (
     OpenRecCaptureIE,
+    OpenRecChannelIE,
+    OpenRecChannelSearchIE,
     OpenRecIE,
     OpenRecMovieIE,
+    OpenRecPlaylistIE,
 )
 from .orf import (
     ORFIPTVIE,

yt_dlp/extractor/bilibili.py

@@ -127,7 +127,7 @@ class BilibiliBaseIE(InfoExtractor):
                     'format_note': ('quality', {format_names.get}),
                     'duration': ('timelength', {float_or_none(scale=1000)}),
                 }),
-                **parse_resolution(format_names.get(play_info.get('quality'))),
+                **parse_resolution(traverse_obj(play_info, ('quality', {format_names.get}))),
             })
         return formats
 
@@ -166,8 +166,58 @@ class BilibiliBaseIE(InfoExtractor):
         params['w_rid'] = hashlib.md5(f'{query}{self._get_wbi_key(video_id)}'.encode()).hexdigest()
         return params
 
-    def _download_playinfo(self, bvid, cid, headers=None, query=None):
+    @staticmethod
-        params = {'bvid': bvid, 'cid': cid, 'fnval': 4048, **(query or {})}
+    @functools.cache
+    def __screen_dimensions():
+        dims, prefs = zip(
+            ((1920, 1080), 18),
+            ((1366, 768), 18),
+            ((1536, 864), 17),
+            ((1280, 720), 8),
+            ((2560, 1440), 7),
+            ((1440, 900), 5),
+            ((1600, 900), 5),
+            strict=True)
+        return random.choices(dims, weights=prefs)[0]
+
+    @property
+    def _dm_params(self):
+        def get_wh(width=1920, height=1080):
+            res0, res1 = width, height
+            rnd = math.floor(114 * random.random())
+            return [2 * res0 + 2 * res1 + 3 * rnd, 4 * res0 - res1 + rnd, rnd]
+
+        def get_of(scroll_top=10, scroll_left=10):
+            res0, res1 = scroll_top, scroll_left
+            rnd = math.floor(514 * random.random())
+            return [3 * res0 + 2 * res1 + rnd, 4 * res0 - 4 * res1 + 2 * rnd, rnd]
+
+        # Source: https://s1.hdslb.com/bfs/seed/jinkela/short/user-fingerprint/bili-user-fingerprint.min.js
+        # function window.__biliUserFp__.queryUserLog
+        # .dm_img_list and .dm_img_inter.ds are more troublesome as they come from mousemove/click events.
+        # Leave them empty for now, since they should allow playing the video without any mousemove/click.
+        return {
+            'dm_img_list': '[]',
+            'dm_img_str': base64.b64encode(
+                ''.join(random.choices(string.printable, k=random.randint(16, 64))).encode())[:-2].decode(),
+            'dm_cover_img_str': base64.b64encode(
+                ''.join(random.choices(string.printable, k=random.randint(32, 128))).encode())[:-2].decode(),
+            # Bilibili expects dm_img_inter to be a compact JSON (without spaces)
+            'dm_img_inter': json.dumps({
+                'ds': [],
+                'wh': get_wh(*self.__screen_dimensions()),
+                'of': get_of(random.randint(0, 100), 0),
+            }, separators=(',', ':')),
+        }
+
+    def _download_playinfo(self, bvid, cid, headers=None, query=None, fatal=True):
+        params = {
+            'bvid': bvid,
+            'cid': cid,
+            'fnval': 4048,
+            **self._dm_params,
+            **(query or {}),
+        }
         if self.is_logged_in:
             params.pop('try_look', None)
         if qn := params.get('qn'):
@@ -175,9 +225,24 @@ class BilibiliBaseIE(InfoExtractor):
         else:
             note = f'Downloading video formats for cid {cid}'
 
-        return self._download_json(
+        playurl_raw = self._download_json(
             'https://api.bilibili.com/x/player/wbi/playurl', bvid,
-            query=self._sign_wbi(params, bvid), headers=headers, note=note)['data']
+            query=self._sign_wbi(params, bvid), headers=headers, note=note)
+        code = traverse_obj(playurl_raw, ('code', {lambda x: x * -1}))
+        if code == 0:
+            return playurl_raw['data']
+        else:
+            msg = join_nonempty(
+                'Unable to download video info', code,
+                traverse_obj(playurl_raw, ('message', {str})),
+                delim=': ')
+            expected = code in (401, 352)
+            if expected:
+                msg += ', please wait and try later'
+            if fatal:
+                raise ExtractorError(msg, expected=expected)
+            else:
+                self.report_warning(msg)
 
     def json2srt(self, json_data):
         srt_data = ''
@@ -298,7 +363,7 @@ class BilibiliBaseIE(InfoExtractor):
                 'title': f'{metainfo.get("title")} - {next(iter(edges.values())).get("title")}',
                 'formats': self.extract_formats(play_info),
                 'description': f'{json.dumps(edges, ensure_ascii=False)}\n{metainfo.get("description", "")}',
-                'duration': float_or_none(play_info.get('timelength'), scale=1000),
+                'duration': traverse_obj(play_info, ('timelength', {float_or_none(scale=1000)})),
                 'subtitles': self.extract_subtitles(video_id, cid),
             }
 
@@ -662,7 +727,10 @@ class BiliBiliIE(BilibiliBaseIE):
         if not self._match_valid_url(urlh.url):
             return self.url_result(urlh.url)
 
-        headers['Referer'] = url
+        headers.update({
+            'Referer': 'https://www.bilibili.com/',
+            'Origin': 'https://www.bilibili.com',
+        })
 
         initial_state = self._search_json(r'window\.__INITIAL_STATE__\s*=', webpage, 'initial state', video_id, default=None)
         if not initial_state:
@@ -758,13 +826,12 @@ class BiliBiliIE(BilibiliBaseIE):
                 duration=traverse_obj(initial_state, ('videoData', 'duration', {int_or_none})),
                 __post_extractor=self.extract_comments(aid))
 
-        play_info = None
+        play_info = traverse_obj(
-        if self.is_logged_in:
+            self._search_json(r'window\.__playinfo__\s*=', webpage, 'play info', video_id, default=None),
-            play_info = traverse_obj(
+            ('data', {dict}))
-                self._search_json(r'window\.__playinfo__\s*=', webpage, 'play info', video_id, default=None),
+        if not self.is_logged_in or not play_info:
-                ('data', {dict}))
+            if dl_play_info := self._download_playinfo(video_id, cid, headers=headers, query={'try_look': 1}, fatal=False):
-        if not play_info:
+                play_info = dl_play_info
-            play_info = self._download_playinfo(video_id, cid, headers=headers, query={'try_look': 1})
         formats = self.extract_formats(play_info)
 
         if video_data.get('is_upower_exclusive'):
@@ -819,13 +886,13 @@ class BiliBiliIE(BilibiliBaseIE):
                     'subtitles': self.extract_subtitles(video_id, cid) if idx == 0 else None,
                     '__post_extractor': self.extract_comments(aid) if idx == 0 else None,
                 } for idx, fragment in enumerate(formats[0]['fragments'])],
-                'duration': float_or_none(play_info.get('timelength'), scale=1000),
+                'duration': traverse_obj(play_info, ('timelength', {float_or_none(scale=1000)})),
             }
 
         return {
             **metainfo,
             'formats': formats,
-            'duration': float_or_none(play_info.get('timelength'), scale=1000),
+            'duration': traverse_obj(play_info, ('timelength', {float_or_none(scale=1000)})),
             'chapters': self._get_chapters(aid, cid),
             'subtitles': self.extract_subtitles(video_id, cid),
             '__post_extractor': self.extract_comments(aid),
@@ -1319,20 +1386,21 @@ class BilibiliSpaceVideoIE(BilibiliSpaceBaseIE):
                 'pn': page_idx + 1,
                 'ps': 30,
                 'tid': 0,
-                'web_location': 1550101,
+                'web_location': '333.1387',
-                'dm_img_list': '[]',
+                'special_type': '',
-                'dm_img_str': base64.b64encode(
+                'index': 0,
-                    ''.join(random.choices(string.printable, k=random.randint(16, 64))).encode())[:-2].decode(),
+                **self._dm_params,
-                'dm_cover_img_str': base64.b64encode(
-                    ''.join(random.choices(string.printable, k=random.randint(32, 128))).encode())[:-2].decode(),
-                'dm_img_inter': '{"ds":[],"wh":[6093,6631,31],"of":[430,760,380]}',
             }
 
             try:
                 response = self._download_json(
                     'https://api.bilibili.com/x/space/wbi/arc/search', playlist_id,
                     query=self._sign_wbi(query, playlist_id),
-                    note=f'Downloading space page {page_idx}', headers={'Referer': url})
+                    note=f'Downloading space page {page_idx}', headers={
+                        'Referer': url,
+                        'Origin': 'https://space.bilibili.com',
+                        'Accept-Language': 'en,zh-CN;q=0.9,zh;q=0.8',
+                    })
             except ExtractorError as e:
                 if isinstance(e.cause, HTTPError) and e.cause.status == 412:
                     raise ExtractorError(
@@ -2030,12 +2098,9 @@ class BiliBiliDynamicIE(InfoExtractor):
 
     def _real_extract(self, url):
         post_id = self._match_id(url)
-        # Without the newer chrome UA, the API will return an error (-352)
         post_data = self._download_json(
             'https://api.bilibili.com/x/polymer/web-dynamic/v1/detail', post_id,
-            query={'id': post_id}, headers={
+            query={'id': post_id})
-                'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/131.0.0.0 Safari/537.36',
-            })
         video_url = traverse_obj(post_data, (
             'data', 'item', (None, 'orig'), 'modules', 'module_dynamic',
             (('major', ('archive', 'pgc')), ('additional', ('reserve', 'common'))),

yt_dlp/extractor/instagram.py

@@ -3,9 +3,11 @@ import hashlib
 import itertools
 import json
 import re
+import urllib.parse
 
 from .common import InfoExtractor
 from ..networking.exceptions import HTTPError
+from ..networking.impersonate import ImpersonateTarget
 from ..utils import (
     ExtractorError,
     bug_reports_message,
@@ -42,15 +44,27 @@ def _id_to_pk(shortcode):
 class InstagramBaseIE(InfoExtractor):
     _API_BASE_URL = 'https://i.instagram.com/api/v1'
     _BASE_URL = 'https://www.instagram.com/'
-    _LOGIN_URL = 'https://www.instagram.com/accounts/login'
     _APP_IDS = {
         'ios': '124024574287414',
         'web': '936619743392459',  # default
     }
+    _AUTH_COOKIE_NAME = 'sessionid'
+    _COOKIE_DOMAINS = (
+        'i.instagram.com',
+        '.i.instagram.com',
+        'www.instagram.com',
+        '.www.instagram.com',
+        'instagram.com',
+        '.instagram.com',
+    )
+
+    @functools.cached_property
+    def _can_impersonate(self):
+        return self._downloader._impersonate_target_available(ImpersonateTarget())
 
     @property
     def _is_logged_in(self):
-        return bool(self._get_cookies(self._BASE_URL).get('sessionid'))
+        return bool(self._get_cookies(self._BASE_URL).get(self._AUTH_COOKIE_NAME))
 
     @functools.cached_property
     def _app_id(self):
@@ -71,6 +85,10 @@ class InstagramBaseIE(InfoExtractor):
             'Accept': '*/*',
         }
 
+    @staticmethod
+    def _is_login_redirect(url):
+        return urllib.parse.urlparse(url).path.startswith('/accounts/login')
+
     def _get_count(self, media, kind, *keys):
         return traverse_obj(
             media, (kind, 'count'), *((f'edge_media_{key}', 'count') for key in keys),
@@ -397,9 +415,11 @@ class InstagramIE(InstagramBaseIE):
 
     def _real_initialize(self):
         if self._is_logged_in:
+            self.write_debug('Found Instagram account cookies')
             return
         if not self._lsd_token:
-            webpage = self._download_webpage(self._BASE_URL, None, 'Setting up session', impersonate=True)
+            webpage = self._download_webpage(
+                self._BASE_URL, None, 'Setting up session', impersonate=self._can_impersonate)
             eqmc = self._search_json(
                 r'<script\b[^>]*\bid="__eqmc"[^>]*>', webpage, 'eqmc JSON', None, default={})
             self._lsd_token = (
@@ -411,15 +431,27 @@ class InstagramIE(InstagramBaseIE):
         media_id = str(_id_to_pk(video_id))
 
         if self._is_logged_in:
-            return self._extract_product(self._download_json(
+            try:
-                f'{self._API_BASE_URL}/media/{media_id}/info/', video_id,
+                return self._extract_product(self._download_json(
-                'Downloading video info', 'Video info extraction failed',
+                    f'{self._API_BASE_URL}/media/{media_id}/info/', video_id,
-                impersonate=self._is_web_app, headers=self._api_headers)['items'][0])
+                    'Downloading video info', 'Video info extraction failed',
+                    impersonate=self._can_impersonate and self._is_web_app,
+                    headers=self._api_headers)['items'][0])
+            except ExtractorError as e:
+                if not (isinstance(e.cause, HTTPError) and self._is_login_redirect(e.cause.response.url)):
+                    raise
+
+            self.report_warning('The provided Instagram account cookies are no longer valid')
+            # XXX: With curl-cffi, the error response may not invalidate the cookie in our jar
+            for domain in self._COOKIE_DOMAINS:
+                self.cookiejar.clear(domain=domain, path='/', name=self._AUTH_COOKIE_NAME)
+            # Re-initialize to set lsd token for logged-out extraction
+            self._real_initialize()
 
         api_check = self._download_json(
             f'{self._API_BASE_URL}/web/get_ruling_for_content/', video_id,
             'Checking post accessibility', errnote=False, fatal=False,
-            impersonate=True, headers=self._api_headers,
+            impersonate=self._can_impersonate, headers=self._api_headers,
             query={'content_type': 'MEDIA', 'target_id': media_id}) or {}
 
         csrf_token = self._get_cookies('https://www.instagram.com').get('csrftoken')
@@ -447,7 +479,7 @@ class InstagramIE(InstagramBaseIE):
                 'server_timestamps': 'true',
                 'variables': json.dumps({'media_id': media_id}, separators=(',', ':')),
                 'doc_id': '27130156389949648',
-            }))
+            })) if self._can_impersonate else None
 
         media = traverse_obj(response, ('data', 'xig_polaris_media', {dict}))
         product_info = traverse_obj(media, ('if_not_gated_logged_out', {dict}))
@@ -465,8 +497,8 @@ class InstagramIE(InstagramBaseIE):
                     'This content is only available for registered users who follow this account')
 
             webpage, urlh = self._download_webpage_handle(
-                f'https://www.instagram.com/p/{video_id}', video_id)
+                f'https://www.instagram.com/p/{video_id}', video_id, impersonate=self._can_impersonate)
-            if urlh.url.startswith(self._LOGIN_URL):
+            if self._is_login_redirect(urlh.url):
                 self.raise_login_required(
                     'The webpage request was redirected to the login page. '
                     'You have exceeded the rate-limit for accessing posts anonymously')
@@ -685,7 +717,8 @@ class InstagramStoryIE(InstagramBaseIE):
         if username == 'highlights' and not story_id:  # story id is only mandatory for highlights
             raise ExtractorError('Input URL is missing a highlight ID', expected=True)
         display_id = story_id or username
-        story_info = self._download_webpage(url, display_id, impersonate=self._is_web_app)
+        story_info = self._download_webpage(
+            url, display_id, impersonate=self._can_impersonate and self._is_web_app)
         user_info = self._search_json(r'"user":', story_info, 'user info', display_id, fatal=False)
         if not user_info:
             self.raise_login_required('This content is unreachable')
@@ -700,8 +733,8 @@ class InstagramStoryIE(InstagramBaseIE):
 
         videos = traverse_obj(self._download_json(
             f'{self._API_BASE_URL}/feed/reels_media/?reel_ids={story_info_url}',
-            display_id, errnote=False, fatal=False, impersonate=self._is_web_app,
+            display_id, errnote=False, fatal=False, headers=self._api_headers,
-            headers=self._api_headers), 'reels')
+            impersonate=self._can_impersonate and self._is_web_app), 'reels')
         if not videos:
             self.raise_login_required('You need to log in to access this content')
         user_info = traverse_obj(videos, (user_id, 'user', {dict})) or {}

yt_dlp/extractor/omnyfm.py

@@ -0,0 +1,272 @@
+import functools
+import itertools
+
+from .common import InfoExtractor
+from ..utils import (
+    OnDemandPagedList,
+    clean_html,
+    filter_dict,
+    float_or_none,
+    int_or_none,
+    parse_duration,
+    parse_iso8601,
+    parse_qs,
+    update_url,
+    url_or_none,
+)
+from ..utils.traversal import (
+    require,
+    traverse_obj,
+    trim_str,
+)
+
+
+class OmnyfmIE(InfoExtractor):
+    _VALID_URL = r'https?://omny\.fm/shows/(?P<uploader_id>[\w-]+)/(?P<id>(?!playlists(?:[/?#"\']|$))[\w-]+)(?:/embed)?(?=[?#"\']|$)'
+    _EMBED_REGEX = [rf'<iframe[^>]+\bsrc\s*=\s*(["\'])(?P<url>{_VALID_URL}[^"\']*)\1']
+    _TESTS = [{
+        'url': 'https://omny.fm/shows/sleep-hub/cannabinoids-and-sleep',
+        'md5': 'e45ec0ce43da757a0be6ca117ec01bdc',
+        'info_dict': {
+            'id': 'cannabinoids-and-sleep',
+            'ext': 'mp3',
+            'title': 'Cannabinoids and Sleep',
+            'categories': 'count:1',
+            'chapters': [
+                {'start_time': 0, 'title': 'Introduction'},
+                {'start_time': 138, 'title': 'Theme: Cannabinoids and Sleep'},
+                {'start_time': 1487, 'title': 'Clinical Tip'},
+                {'start_time': 1635, 'title': 'Pick of the Month'},
+                {'start_time': 1795, 'title': 'What\'s Coming Up?'},
+            ],
+            'description': 'md5:c0fd2d29f3148382d344cfbd012fb00d',
+            'duration': 1840.274,
+            'episode': 'Episode 48',
+            'episode_number': 48,
+            'modified_date': r're:\d{8}',
+            'modified_timestamp': int,
+            'tags': 'count:6',
+            'thumbnail': r're:https?://www\.omnycontent\.com/.+',
+            'timestamp': 1574013600,
+            'upload_date': '20191117',
+            'uploader': 'Sleep Talk',
+            'uploader_id': 'sleep-hub',
+        },
+    }, {
+        'url': 'https://omny.fm/shows/the-origin-of-things/a-song-of-hope/embed',
+        'md5': 'd7600ef33e3f139ff1bb8946f3651b15',
+        'info_dict': {
+            'id': 'a-song-of-hope',
+            'ext': 'mp3',
+            'title': 'A song of hope',
+            'categories': 'count:3',
+            'description': 'md5:f8e710765c341a48cfda8dacd428f56d',
+            'duration': 478.955,
+            'episode': 'Episode 17',
+            'episode_number': 17,
+            'modified_date': r're:\d{8}',
+            'modified_timestamp': int,
+            'season': 'Season 3',
+            'season_number': 3,
+            'tags': 'count:27',
+            'thumbnail': r're:https?://www\.omnycontent\.com/.+',
+            'timestamp': 1679445000,
+            'upload_date': '20230322',
+            'uploader': 'The Origin Of Things',
+            'uploader_id': 'the-origin-of-things',
+        },
+    }]
+    _WEBPAGE_TESTS = [{
+        'url': 'https://www.asahi.com/special/podcasts/item/?itemid=311a5f48-ad71-4548-b1f2-af5e00747fbc',
+        'md5': '4c788bf03323734524a7c0f98d9956ed',
+        'info_dict': {
+            'id': 'sdgs-271',
+            'ext': 'mp3',
+            'title': '「どこかのだれかの人生のにおいがする」 SDGsを音声番組で身近に #271',
+            'categories': 'count:5',
+            'description': 'md5:fa086ecce764d81c51648a82d0fe4850',
+            'duration': 1870.524,
+            'episode': 'Episode 271',
+            'episode_number': 271,
+            'modified_date': r're:\d{8}',
+            'modified_timestamp': int,
+            'season': 'Season 1',
+            'season_number': 1,
+            'tags': 'count:4',
+            'thumbnail': r're:https?://www\.omnycontent\.com/.+',
+            'timestamp': 1670266800,
+            'upload_date': '20221205',
+            'uploader': '朝日新聞ポッドキャスト',
+            'uploader_id': 'asahi',
+            'webpage_url': 'https://omny.fm/shows/asahi/sdgs-271',
+        },
+    }]
+
+    def _real_extract(self, url):
+        uploader_id, audio_id = self._match_valid_url(url).group('uploader_id', 'id')
+        webpage = self._download_webpage(url, audio_id)
+        nextjs_data = self._search_nextjs_data(webpage, audio_id)
+        clip = traverse_obj(nextjs_data, ('props', 'pageProps', 'clip', {dict}))
+
+        return {
+            'id': audio_id,
+            'section_start': traverse_obj(url, ({parse_qs}, 't', -1, {parse_duration})),
+            'uploader_id': uploader_id,
+            'vcodec': 'none',
+            **traverse_obj(clip, {
+                'title': ('Title', {clean_html}, filter),
+                'chapters': ('Chapters', lambda _, v: parse_duration(v['Position']) is not None, {
+                    'title': ('Name', {clean_html}, filter),
+                    'start_time': ('Position', {parse_duration}),
+                }),
+                'description': ('Description', {clean_html}, filter),
+                'duration': ('DurationSeconds', {float_or_none}),
+                'episode_number': ('Episode', {int_or_none}),
+                'filesize': ('PublishedAudioSizeInBytes', {int_or_none}),
+                'modified_timestamp': ('ModifiedAtUtc', {parse_iso8601}),
+                'season_number': ('Season', {int_or_none}),
+                'tags': ('Tags', ..., {clean_html}, filter, all, filter),
+                'thumbnail': ('ImageUrl', {update_url(query=None)}),
+                'timestamp': ('PublishedUtc', {parse_iso8601}),
+                'url': ('AudioUrl', {url_or_none}, {require('audio URL')}),
+                'webpage_url': ('PublishedUrl', {url_or_none}),
+            }),
+            **traverse_obj(clip, ('Program', {
+                'categories': ('Categories', ..., {clean_html}, filter, all, filter),
+                'uploader': ('Name', {clean_html}, filter),
+            })),
+        }
+
+
+class OmnyfmPlaylistBaseIE(InfoExtractor):
+    _API_BASE = 'https://api.omny.fm'
+    _BASE_URL = 'https://omny.fm/shows'
+    _PAGE_SIZE = 100
+
+    def _yield_clips(self, clips, uploader_id):
+        for audio_id in traverse_obj(clips, (
+            'Clips', ..., 'Slug', {str},
+        )):
+            yield self.url_result(
+                f'{self._BASE_URL}/{uploader_id}/{audio_id}', OmnyfmIE)
+
+
+class OmnyfmPlaylistIE(OmnyfmPlaylistBaseIE):
+    _VALID_URL = r'https?://omny\.fm/shows/(?P<uploader_id>[\w-]+)/playlists(?:/(?P<id>[\w-]+))?(?:/embed)?/?(?=[?#"\']|$)'
+    _EMBED_REGEX = [fr'<iframe[^>]+\bsrc=(["\'])(?P<url>{_VALID_URL}[^"\']*)\1']
+    _TESTS = [{
+        'url': 'https://omny.fm/shows/sleep-hub/playlists/sleep-talk',
+        'info_dict': {
+            'id': 'sleep-talk',
+            'title': 'Sleep Talk - Talking all things sleep',
+            'description': 'md5:c1d7e5bf32100a432307d2d32c4ab74a',
+            'thumbnail': r're:https?://www\.omnycontent\.com/.+',
+        },
+        'playlist_mincount': 79,
+    }, {
+        'url': 'https://omny.fm/shows/bayfm-program03/playlists',
+        'info_dict': {
+            'id': 'bayfm-program03',
+        },
+        'playlist_count': 4,
+    }]
+    _WEBPAGE_TESTS = [{
+        'url': 'https://www.asahi.com/articles/ASP763WDKP4JDIFI002.html',
+        'info_dict': {
+            'id': 'podcast',
+            'title': 'ニュースの現場から',
+            'description': 'md5:ed1f78462ebed09258ca31b1da5ff640',
+            'thumbnail': r're:https?://www\.omnycontent\.com/.+',
+            'webpage_url': 'https://omny.fm/shows/asahi/playlists/podcast',
+        },
+        'playlist_mincount': 2517,
+    }]
+
+    def _entries(self, uploader_id, playlist_id):
+        clip_id = None
+
+        for page in itertools.count(1):
+            clips = self._download_json(
+                f'{self._API_BASE}/programs/{uploader_id}/playlists/{playlist_id}/clips',
+                playlist_id, f'Downloading page {page}', query=filter_dict({
+                    'clipId': clip_id,
+                    'direction': 'AfterExclusive',
+                    'pageSize': self._PAGE_SIZE,
+                }))
+
+            yield from self._yield_clips(clips, uploader_id)
+
+            if not clips.get('NextClipsAvailable'):
+                break
+
+            clip_id = traverse_obj(clips, ('Clips', -1, 'Id', {str}))
+            if not clip_id:
+                break
+
+    def _real_extract(self, url):
+        uploader_id, playlist_id = self._match_valid_url(url).group('uploader_id', 'id')
+        webpage = self._download_webpage(url, playlist_id or uploader_id)
+        nextjs_data = self._search_nextjs_data(webpage, playlist_id or uploader_id)
+        page_props = traverse_obj(nextjs_data, ('props', 'pageProps', {dict}))
+
+        if not playlist_id:
+            entries = [self.url_result(
+                f'{self._BASE_URL}/{uploader_id}/playlists/{playlist_id}', OmnyfmPlaylistIE,
+            ) for playlist_id in traverse_obj(page_props, (
+                'playlistsWithClips', ..., 'playlist', 'Slug', {str},
+            ))]
+
+            return self.playlist_result(entries, uploader_id)
+
+        return self.playlist_result(
+            self._entries(uploader_id, playlist_id), playlist_id,
+            **traverse_obj(page_props, ('playlist', {
+                'title': ('Title', {clean_html}, filter),
+                'description': ('Description', {clean_html}, filter),
+                'thumbnail': ('ArtworkUrl', {update_url(query=None)}),
+                'webpage_url': ('EmbedUrl', {url_or_none}, {trim_str(end='/embed')}),
+            })))
+
+
+class OmnyfmShowIE(OmnyfmPlaylistBaseIE):
+    _VALID_URL = r'https?://omny\.fm/shows/(?P<id>[\w-]+)/?(?:[?#]|$)'
+    _TESTS = [{
+        'url': 'https://omny.fm/shows/the-origin-of-things',
+        'info_dict': {
+            'id': 'the-origin-of-things',
+            'title': 'The Origin Of Things',
+            'description': 'md5:52b7fba08201d050639c78ea88cc782e',
+            'thumbnail': r're:https?://www\.omnycontent\.com/.+',
+        },
+        'playlist_mincount': 75,
+    }]
+
+    def _fetch_page(self, uploader_id, organization_id, program_id, page):
+        clips = self._download_json(
+            f'{self._API_BASE}/orgs/{organization_id}/programs/{program_id}/clips',
+            uploader_id, f'Downloading page {page + 1}', query={
+                'cursor': page,
+                'pageSize': self._PAGE_SIZE,
+            })
+
+        yield from self._yield_clips(clips, uploader_id)
+
+    def _real_extract(self, url):
+        uploader_id = self._match_id(url)
+        webpage = self._download_webpage(url, uploader_id)
+        nextjs_data = self._search_nextjs_data(webpage, uploader_id)
+        program = traverse_obj(nextjs_data, ('props', 'pageProps', 'program', {dict}))
+
+        organization_id = traverse_obj(program, (
+            'OrganizationId', {str}, {require('organization ID')}))
+        program_id = traverse_obj(program, ('Id', {str}, {require('program ID')}))
+        entries = OnDemandPagedList(
+            functools.partial(self._fetch_page, uploader_id, organization_id, program_id), self._PAGE_SIZE)
+
+        return self.playlist_result(
+            entries, uploader_id,
+            **traverse_obj(program, {
+                'title': ('Name', {clean_html}, filter),
+                'description': ('Description', {clean_html}, filter),
+                'thumbnail': ('ArtworkUrl', {update_url(query=None)}),
+            }))

yt_dlp/extractor/openrec.py

@@ -1,153 +1,728 @@
+import collections
+import datetime as dt
+import functools
+import itertools
+import json
+import math
+import time
+import urllib.parse
+import xml.etree.ElementTree
+
 from .common import InfoExtractor
 from ..utils import (
     ExtractorError,
-    get_first,
+    InAdvancePagedList,
+    clean_html,
+    extract_attributes,
+    filter_dict,
     int_or_none,
+    join_nonempty,
+    parse_iso8601,
+    parse_qs,
+    str_or_none,
+    update_url_query,
+    url_or_none,
+    urljoin,
+)
+from ..utils.traversal import (
+    find_element,
+    find_elements,
+    require,
     traverse_obj,
-    try_get,
-    unified_strdate,
-    unified_timestamp,
 )
 
 
 class OpenRecBaseIE(InfoExtractor):
-    _M3U8_HEADERS = {'Referer': 'https://www.openrec.tv/'}
+    _API_BASE = 'https://apiv5.mellow-fan.com/api/v5'
+    _BASE_URL = 'https://www.mellow-fan.com'
+    _HEADERS = {'Referer': f'{_BASE_URL}/'}
+    _NETRC_MACHINE = 'mellowfan'
+    _PUBLIC_API_BASE = 'https://public.mellow-fan.com/external/api/v5'
+
+    def _perform_login(self, username, password):
+        if self._get_cookies(self._BASE_URL).get('access-token'):
+            return
+
+        login = self._download_json(
+            f'{self._BASE_URL}/apiv5/email/login',
+            None, 'Logging in', headers={
+                'Content-Type': 'application/json',
+            }, data=json.dumps({
+                'email': username,
+                'password': password,
+            }).encode())
+        if traverse_obj(login, ('status', {int_or_none})) != 0:
+            err_msg = traverse_obj(login, ('message', {clean_html}, filter))
+            raise ExtractorError(err_msg or 'Failed to log in', expected=True)
+
+    def _real_initialize(self):
+        cookies = self._get_cookies(self._BASE_URL)
+        self._api_headers = traverse_obj(cookies, {
+            'access-token': ('access_token', 'value', {str}, filter),
+            'random': ('random', 'value', {str}, filter),
+            'token': ('token', 'value', {str}, filter),
+            'uuid': ('uuid', 'value', {str}, filter),
+        })
 
     def _extract_pagestore(self, webpage, video_id):
-        return self._parse_json(
+        start = r'window\.pageStore\s*='
-            self._search_regex(r'(?m)window\.pageStore\s*=\s*(\{.+?\});$', webpage, 'window.pageStore'), video_id)
 
-    def _expand_media(self, video_id, media):
+        if store := self._search_regex(
-        for name, m3u8_url in (media or {}).items():
+                rf'{start}\s*JSON\.parse\s*\(\s*decodeURIComponent'
-            if not m3u8_url:
+                r'\s*\(\s*(?P<q>["\'])(?P<json>.*?)(?P=q)\s*\)\s*\)',
-                continue
+                webpage, 'encoded window pagestore', group='json', default=None,
-            yield from self._extract_m3u8_formats(
+        ):
-                m3u8_url, video_id, ext='mp4', m3u8_id=name, headers=self._M3U8_HEADERS)
+            return self._parse_json(store, video_id, transform_source=urllib.parse.unquote)
+        return self._search_json(start, webpage, 'window pagestore', video_id)
 
-    def _extract_movie(self, webpage, video_id, name, is_live):
+    def _call_api(self, path, item_id):
-        window_stores = self._extract_pagestore(webpage, video_id)
+        return self._download_json(
-        movie_stores = [
+            f'{self._API_BASE}/{path}', item_id,
-            # extract all three important data (most of data are duplicated each other, but slightly different!)
+            headers=self._api_headers, expected_status=401)
-            traverse_obj(window_stores, ('v8', 'state', 'movie'), expected_type=dict),
-            traverse_obj(window_stores, ('v8', 'movie'), expected_type=dict),
-            traverse_obj(window_stores, 'movieStore', expected_type=dict),
-        ]
-        if not any(movie_stores):
-            raise ExtractorError(f'Failed to extract {name} info')
 
-        formats = list(self._expand_media(video_id, get_first(movie_stores, 'media')))
+    def _parse_openrec_metadata(self, page_store, video_id):
-        if not formats:
+        info = traverse_obj(page_store, ('v8', 'movie', {dict}))
-            # archived livestreams or subscriber-only videos
-            cookies = self._get_cookies('https://www.openrec.tv/')
-            detail = self._download_json(
-                f'https://apiv5.openrec.tv/api/v5/movies/{video_id}/detail', video_id,
-                headers={
-                    'Origin': 'https://www.openrec.tv',
-                    'Referer': 'https://www.openrec.tv/',
-                    'access-token': try_get(cookies, lambda x: x.get('access_token').value),
-                    'uuid': try_get(cookies, lambda x: x.get('uuid').value),
-                })
-            new_media = traverse_obj(detail, ('data', 'items', ..., 'media'), get_all=False)
-            formats = list(self._expand_media(video_id, new_media))
-            is_live = False
 
-        return {
+        target_members = traverse_obj(info, (
+            'targetMembers', ..., 'type', {str}, filter, any))
+        needs_subscription = target_members == 'subscription'
+        needs_auth = target_members == 'ppv'
+
+        me = self._call_api('users/me', video_id)
+        needs_premium = traverse_obj(info, (
+            'publicType', {str}, filter)) == 'premium'
+        is_premium = traverse_obj(me, (
+            'data', 'items', ..., 'is_premium', {bool}, any)) or False
+
+        detail = self._call_api(f'movies/{video_id}/detail', video_id)
+        is_member = traverse_obj(detail, (
+            'data', 'items', ..., 'membership', 'is_active', {bool}, any)) or False
+        has_ppv = traverse_obj(detail, (
+            'data', 'items', ..., 'ppv_ticket_products', ..., {dict}, any)) or False
+
+        need = None
+        if needs_premium and not is_premium:
+            need = 'premium membership'
+        elif needs_subscription and not is_member:
+            need = 'channel subscription'
+        elif needs_auth and not has_ppv:
+            need = 'PPV purchase'
+
+        if need:
+            self.raise_login_required(
+                f'This content requires a {need}', metadata_available=True)
+
+        return info, detail, {
             'id': video_id,
-            'title': get_first(movie_stores, 'title'),
+            'availability': self._availability(
-            'description': get_first(movie_stores, 'introduction'),
+                needs_premium=needs_premium,
-            'thumbnail': get_first(movie_stores, 'thumbnailUrl'),
+                needs_subscription=needs_subscription,
-            'formats': formats,
+                needs_auth=needs_auth,
-            'uploader': get_first(movie_stores, ('channel', 'user', 'name')),
+            ) or 'public',
-            'uploader_id': get_first(movie_stores, ('channel', 'user', 'id')),
+            'http_headers': self._HEADERS,
-            'timestamp': int_or_none(get_first(movie_stores, ['publishedAt', 'time']), scale=1000) or unified_timestamp(get_first(movie_stores, 'publishedAt')),
+            'tags': traverse_obj(page_store, (
-            'is_live': is_live,
+                'movieStore', 'tags', ..., {clean_html}, filter, all, filter)),
-            'http_headers': self._M3U8_HEADERS,
+            **traverse_obj(info, {
+                'title': ('title', {clean_html}, filter),
+                'cast': ('casts', ..., 'name', {clean_html}, filter, all, filter),
+                'categories': ('game', 'title', {clean_html}, filter, all, filter),
+                'description': ('introduction', {clean_html}, filter),
+                'duration': ('playTime', 'value', {int_or_none(scale=1000)}),
+                'thumbnail': (('lThumbnailUrl', 'thumbnailUrl'), {url_or_none}, any),
+                'timestamp': ('startedAt', 'time', {int_or_none(scale=1000)}),
+                'view_count': ('totalViews', {int_or_none}),
+            }),
+            **traverse_obj(info, ('channel', 'user', {
+                'channel_follower_count': ('followers', {int_or_none}),
+                'channel_id': ('id', {str_or_none}),
+                'channel': ('name', {clean_html}, filter),
+                'channel_is_verified': ('isOfficial', {bool}),
+            })),
         }
 
 
 class OpenRecIE(OpenRecBaseIE):
-    IE_NAME = 'openrec'
+    IE_NAME = 'mellowfan'
-    _VALID_URL = r'https?://(?:www\.)?openrec\.tv/live/(?P<id>[^/?#]+)'
+    IE_DESC = 'mellow-fan'
+
+    _VALID_URL = r'https?://(?:www\.)?(?:mellow-fan\.com|openrec\.tv)/(?:m/)?live/(?P<id>[^/?#]+)'
     _TESTS = [{
-        'url': 'https://www.openrec.tv/live/2p8v31qe4zy',
+        'url': 'https://www.openrec.tv/live/e2zwj0mp6ro',
-        'only_matching': True,
+        'info_dict': {
+            'id': 'e2zwj0mp6ro',
+            'ext': 'mp4',
+            'title': '収束',
+            'availability': 'public',
+            'categories': ['雑談'],
+            'channel': 'おおえのたかゆき',
+            'channel_follower_count': int,
+            'channel_id': 'oekaki',
+            'channel_is_verified': True,
+            'comment_count': int,
+            'description': 'md5:62260f3060b40187282f3213bcd97abd',
+            'duration': 14257,
+            'live_status': 'was_live',
+            'tags': ['雑談'],
+            'thumbnail': r're:https?://.+',
+            'timestamp': 1685271819,
+            'upload_date': '20230528',
+            'view_count': int,
+        },
+        'skip': '404 Not Found',
     }, {
-        'url': 'https://www.openrec.tv/live/wez93eqvjzl',
+        # SP
-        'only_matching': True,
+        'url': 'https://www.mellow-fan.com/live/2p8vv29438y',
+        'info_dict': {
+            'id': '2p8vv29438y',
+            'ext': 'mp4',
+            'title': 'それいけ加藤純一探検隊！ 〜南の孤島で希少生物を探せスペシャル〜 (OPENRECプレミアム会員限定)',
+            'availability': 'premium_only',
+            'categories': ['雑談'],
+            'channel': '加藤 純一',
+            'channel_follower_count': int,
+            'channel_id': 'junichi_kato_channel',
+            'channel_is_verified': True,
+            'comment_count': int,
+            'description': 'md5:693bc0c838ff6080f9ca00d41dc5e840',
+            'duration': 24336,
+            'live_status': 'was_live',
+            'release_date': '20251008',
+            'release_timestamp': 1759892400,
+            'thumbnail': r're:https?://.+',
+            'timestamp': 1759891802,
+            'upload_date': '20251008',
+            'view_count': int,
+        },
+    }, {
+        # Members only
+        'url': 'https://www.mellow-fan.com/live/kdr7nldqgzj',
+        'info_dict': {
+            'id': 'kdr7nldqgzj',
+            'ext': 'mp4',
+            'title': '【ゲーム実況生配信】小森結梨のひきこもりゲーム部屋#20[ゲスト:梅澤めぐ]',
+            'availability': 'subscriber_only',
+            'categories': ['Human Fall Flat'],
+            'channel': 'セカンドショットGAME部',
+            'channel_follower_count': int,
+            'channel_id': 'secondshot_games',
+            'channel_is_verified': True,
+            'chapters': [
+                {'start_time': 0, 'title': 'バイオハザード RE：2', 'end_time': 443},
+                {'start_time': 443, 'title': 'ぷにゃん', 'end_time': 3215},
+                {'start_time': 3215, 'title': 'Human Fall Flat', 'end_time': 7163},
+            ],
+            'comment_count': int,
+            'description': 'md5:b77fac6fb0ad86048dc6377ec2c22646',
+            'duration': 7163,
+            'live_status': 'was_live',
+            'release_date': '20260220',
+            'release_timestamp': 1771581300,
+            'tags': ['声優'],
+            'thumbnail': r're:https?://.+',
+            'timestamp': 1771581322,
+            'upload_date': '20260220',
+            'view_count': int,
+        },
+        'skip': 'Subscribers only',
+    }, {
+        # PPV
+        'url': 'https://www.mellow-fan.com/live/e5rk93xn1zv',
+        'info_dict': {
+            'id': 'e5rk93xn1zv',
+            'ext': 'mp4',
+            'title': '三川華月生誕パーティー2026 【ゲスト：幸村恵理/北原沙弥香】',
+            'availability': 'needs_auth',
+            'categories': ['雑談'],
+            'channel': '三川華月の開店！はるちゃん食堂',
+            'channel_follower_count': int,
+            'channel_id': 'haruna_harusyoku',
+            'channel_is_verified': True,
+            'description': 'md5:912d0c2d60d343e28300fc18bd491d6c',
+            'duration': 6451,
+            'live_status': 'was_live',
+            'release_date': '20260222',
+            'release_timestamp': 1771758000,
+            'thumbnail': r're:https?://.+',
+            'timestamp': 1771758062,
+            'upload_date': '20260222',
+            'view_count': int,
+        },
+        'skip': 'Paid video',
     }]
 
+    @staticmethod
+    def _json2xml(subs, started_at):
+        def filter_valid(items):
+            yield from traverse_obj(items, (
+                lambda _, v: clean_html(v['message']) and not v['stamp']))
+
+        total = collections.Counter(
+            traverse_obj(subs, (..., 'posted_at', {parse_iso8601})))
+        order = collections.defaultdict(int)
+
+        root = xml.etree.ElementTree.Element('packet')
+        for i, s in enumerate(filter_valid(subs), 1):
+            posted_at = traverse_obj(s, ('posted_at', {parse_iso8601}))
+            offset = order[posted_at] / total[posted_at] + 1
+            order[posted_at] += 1
+            vpos = int_or_none((posted_at - started_at + offset) * 100)
+
+            xml.etree.ElementTree.SubElement(
+                root, 'chat', filter_dict({
+                    **traverse_obj(s, ('user', {
+                        'premium': ('is_premium', {bool}, {lambda x: '1' if x else '0'}),
+                        'user_id': ('id', {str_or_none}),
+                        'name': ('nickname', {str}, filter),
+                    })),
+                    'no': str(i),
+                    'vpos': str_or_none(vpos),
+                    'date': str_or_none(posted_at),
+                }),
+            ).text = traverse_obj(s, ('message', {clean_html}, filter))
+        xml.etree.ElementTree.indent(root, space='  ')
+
+        return xml.etree.ElementTree.tostring(
+            root, encoding='utf-8', xml_declaration=True).decode()
+
+    def _get_subtitles(self, duration, started_at, video_id):
+        ended_at = started_at + duration
+        timestamp = started_at
+
+        subs = []
+        for page in itertools.count(1):
+            created_at = dt.datetime.fromtimestamp(
+                timestamp, dt.timezone(dt.timedelta(hours=9))).strftime('%Y-%m-%dT%H:%M:%S%z')
+
+            chats = self._download_json(
+                f'{self._PUBLIC_API_BASE}/movies/{video_id}/chats',
+                video_id, f'Downloading chats page {page}', query={
+                    'from_created_at': created_at,
+                    'is_including_system_message': 'true',
+                })
+            if not chats:
+                break
+
+            subs.extend(chats)
+            last_posted_at = traverse_obj(chats, (-1, 'posted_at', {parse_iso8601}))
+            if ended_at < last_posted_at:
+                break
+
+            timestamp = last_posted_at + 1
+            time.sleep(0.1)
+
+        return {
+            'chats': [{
+                'data': json.dumps(subs, indent=2, ensure_ascii=False),
+                'ext': 'json',
+            }, {
+                'data': self._json2xml(subs, started_at),
+                'ext': 'xml',
+            }],
+        }
+
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        webpage = self._download_webpage(f'https://www.openrec.tv/live/{video_id}', video_id)
+        webpage = self._download_webpage(url, video_id, expected_status=404)
+        page_store = self._extract_pagestore(webpage, video_id)
+        if traverse_obj(page_store, ('movieStore', 'notFound', {bool})):
+            raise ExtractorError('This video in no longer available', expected=True)
 
-        return self._extract_movie(webpage, video_id, 'live', True)
+        info, detail, metadata = self._parse_openrec_metadata(page_store, video_id)
+        live_status = {
+            'ARCHIVE': 'was_live',
+            'COMING_UP': 'is_upcoming',
+            'LIVE_STREAMING': 'is_live',
+            'UPLOADED': 'not_live',
+        }.get(info.get('onAirStatus'))
+
+        release_timestamp = traverse_obj(page_store, ('movieStore', 'willStartAt', {parse_iso8601}))
+        if live_status == 'is_upcoming':
+            if release_timestamp is not None:
+                start_time = dt.datetime.fromtimestamp(
+                    release_timestamp, dt.timezone.utc,
+                ).astimezone().strftime('%Y-%m-%d %H:%M:%S %Z')
+                msg = f'This livestream is scheduled to start at {start_time}'
+            else:
+                msg = 'This livestream has not yet started'
+
+            self.raise_no_formats(msg, expected=True)
+
+            return {
+                'id': video_id,
+                'live_status': live_status,
+                'release_timestamp': release_timestamp,
+            }
+
+        duration = metadata['duration']
+        started_at = metadata['timestamp']
+
+        chapters = []
+        for chapter in traverse_obj(info, (
+            'chapters', lambda _, v: int_or_none(v['chapterAt']['time']),
+        )):
+            chapter_at = traverse_obj(chapter, ('chapterAt', 'time', {int_or_none(scale=1000)}))
+            chapters.append({
+                'start_time': chapter_at - started_at,
+                'title': traverse_obj(chapter, ('title', {clean_html}, filter)),
+            })
+
+        formats = []
+        is_dvr = live_status == 'is_live' and self.get_param('live_from_start')
+        media_keys = ('url_dvr', 'url_dvr_audio') if is_dvr else ('url', 'url_audio')
+        for m3u8_url in traverse_obj(detail, (
+            'data', 'items', ...,
+            ('media', 'subs_trial_media'), media_keys, {url_or_none},
+        )):
+            fmts = self._extract_m3u8_formats(
+                m3u8_url, video_id, 'mp4', headers=self._HEADERS)
+            for fmt in fmts:
+                if is_dvr:
+                    fmt.setdefault('downloader_options', {}).update({'ffmpeg_args': ['-live_start_index', '0']})
+                    fmt['is_from_start'] = True
+            formats.extend(fmts)
+
+        return {
+            'chapters': chapters or None,
+            'comment_count': traverse_obj(page_store, (
+                'commentStore', 'commentCount', {int_or_none})),
+            'formats': formats,
+            'live_status': live_status,
+            'release_timestamp': release_timestamp,
+            'subtitles': self.extract_subtitles(duration, started_at, video_id),
+            **metadata,
+        }
 
 
 class OpenRecCaptureIE(OpenRecBaseIE):
-    IE_NAME = 'openrec:capture'
+    IE_NAME = 'mellowfan:capture'
-    _VALID_URL = r'https?://(?:www\.)?openrec\.tv/capture/(?P<id>[^/?#]+)'
+
+    _VALID_URL = r'https?://(?:www\.)?(?:mellow-fan\.com|openrec\.tv)/(?:m/)?capture/(?P<id>[^/?#]+)'
     _TESTS = [{
-        'url': 'https://www.openrec.tv/capture/l9nk2x4gn14',
+        'url': 'https://www.mellow-fan.com/capture/l2q00vxl8q8',
-        'only_matching': True,
-    }, {
-        'url': 'https://www.openrec.tv/capture/mldjr82p7qk',
         'info_dict': {
-            'id': 'mldjr82p7qk',
+            'id': 'l2q00vxl8q8',
-            'title': 'たいじの恥ずかしい英語力',
+            'ext': 'mp4',
-            'uploader': 'たいちゃんねる',
+            'title': '？？？？',
-            'uploader_id': 'Yaritaiji',
+            'channel': '布団ちゃん',
-            'upload_date': '20210803',
+            'channel_id': 'indegnasen',
+            'duration': 89,
+            'thumbnail': r're:https?://.+',
+            'timestamp': 1637589871,
+            'upload_date': '20211122',
+        },
+    }, {
+        'url': 'https://www.mellow-fan.com/capture/9pdz9334vng',
+        'info_dict': {
+            'id': '9pdz9334vng',
+            'ext': 'mp4',
+            'title': 'オプレの現実',
+            'channel': 'ゆゆうた&みゃこの泥沼バラエティ',
+            'channel_id': 'doronuma-variety',
+            'duration': 64,
+            'thumbnail': r're:https?://.+',
+            'timestamp': 1677585253,
+            'upload_date': '20230228',
         },
     }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        webpage = self._download_webpage(f'https://www.openrec.tv/capture/{video_id}', video_id)
+        webpage = self._download_webpage(url, video_id)
+        page_store = self._extract_pagestore(webpage, video_id)
 
-        window_stores = self._extract_pagestore(webpage, video_id)
+        capture = page_store['capture']
-        movie_store = window_stores.get('movie')
+        m3u8_url = traverse_obj(capture, ('source', {url_or_none}))
-
-        capture_data = window_stores.get('capture')
-        if not capture_data:
-            raise ExtractorError('Cannot extract title')
-
-        formats = self._extract_m3u8_formats(
-            capture_data.get('source'), video_id, ext='mp4', headers=self._M3U8_HEADERS)
 
         return {
             'id': video_id,
-            'title': capture_data.get('title'),
+            'formats': self._extract_m3u8_formats(
-            'thumbnail': capture_data.get('thumbnailUrl'),
+                m3u8_url, video_id, 'mp4', headers=self._HEADERS),
-            'formats': formats,
+            'http_headers': self._HEADERS,
-            'timestamp': unified_timestamp(traverse_obj(movie_store, 'createdAt', expected_type=str)),
+            **traverse_obj(page_store, ('movie', 'channel', {
-            'uploader': traverse_obj(movie_store, ('channel', 'name'), expected_type=str),
+                'channel': ('name', {clean_html}, filter),
-            'uploader_id': traverse_obj(movie_store, ('channel', 'id'), expected_type=str),
+                'channel_id': ('id', {str}),
-            'upload_date': unified_strdate(capture_data.get('createdAt')),
+            })),
-            'http_headers': self._M3U8_HEADERS,
+            **traverse_obj(capture, {
+                'title': ('title', {clean_html}, filter),
+                'duration': ({lambda x: int_or_none(x['endTime']) - int_or_none(x['startTime'])}),
+                'thumbnail': ('thumbnailUrl', {url_or_none}),
+                'timestamp': ('publishedAt', {parse_iso8601}),
+            }),
         }
 
 
 class OpenRecMovieIE(OpenRecBaseIE):
-    IE_NAME = 'openrec:movie'
+    IE_NAME = 'mellowfan:movie'
-    _VALID_URL = r'https?://(?:www\.)?openrec\.tv/movie/(?P<id>[^/?#]+)'
+
+    _VALID_URL = r'https?://(?:www\.)?(?:mellow-fan\.com|openrec\.tv)/(?:m/)?movie/(?P<id>[^/?#]+)'
     _TESTS = [{
-        'url': 'https://www.openrec.tv/movie/nqz5xl5km8v',
+        'url': 'https://www.mellow-fan.com/movie/e5rk9k4o6zv',
         'info_dict': {
-            'id': 'nqz5xl5km8v',
+            'id': 'e5rk9k4o6zv',
-            'title': '限定コミュニティ(Discord)参加方法ご説明動画',
+            'ext': 'mp4',
-            'description': 'md5:ebd563e5f5b060cda2f02bf26b14d87f',
+            'title': 'みゃこRaMuの企画会議#3 ～2人のやってみたいこと～',
-            'thumbnail': r're:https://.+',
+            'availability': 'public',
-            'uploader': 'タイキとカズヒロ',
+            'categories': ['雑談'],
-            'uploader_id': 'taiki_to_kazuhiro',
+            'channel': 'みゃことRaMuの何して遊ぶ？',
-            'timestamp': 1638856800,
+            'channel_follower_count': int,
+            'channel_id': 'myakoramu',
+            'channel_is_verified': True,
+            'description': 'md5:90924fd73356ebd574bec8d761d4fa62',
+            'duration': 771,
+            'tags': ['雑談'],
+            'thumbnail': r're:https?://.+',
+            'view_count': int,
         },
     }, {
-        'url': 'https://www.openrec.tv/movie/2p8vvex548y?playlist_id=98brq96vvsgn2nd',
+        # Members only
-        'only_matching': True,
+        'url': 'https://www.mellow-fan.com/movie/n9ze6q3eo84',
+        'info_dict': {
+            'id': 'n9ze6q3eo84',
+            'ext': 'mp4',
+            'title': '大西亜玖璃・高尾奏音のあぐのんる～むらぼ♪第123回傑作選vol.3【高画質・完全版】',
+            'availability': 'subscriber_only',
+            'categories': ['雑談'],
+            'channel': 'セカンドショットGAME部',
+            'channel_follower_count': int,
+            'channel_id': 'secondshot_games',
+            'channel_is_verified': True,
+            'description': 'md5:b7ab5ddd71ba5edc1141cb427af7a9c6',
+            'duration': 1800,
+            'tags': ['声優'],
+            'thumbnail': r're:https?://.+',
+            'view_count': int,
+        },
+        'skip': 'Subscribers only',
+    }, {
+        # PPV
+        'url': 'https://www.mellow-fan.com/movie/em8xvd4ljr2',
+        'info_dict': {
+            'id': 'em8xvd4ljr2',
+            'ext': 'mp4',
+            'title': '【PPV購入特典映像】KAWAII LAB. SESSION in OKINAWA',
+            'availability': 'needs_auth',
+            'categories': ['ミュージック'],
+            'channel': 'KAWAII LAB.チャンネル',
+            'channel_follower_count': int,
+            'channel_id': 'KAWAIILAB',
+            'channel_is_verified': True,
+            'description': 'md5:e9f67d8648d3cbe35df1dd689afc29a8',
+            'duration': 2002,
+            'tags': ['アイドル'],
+            'thumbnail': r're:https?://.+',
+            'view_count': int,
+        },
+        'skip': 'Paid video',
     }]
 
     def _real_extract(self, url):
         video_id = self._match_id(url)
-        webpage = self._download_webpage(f'https://www.openrec.tv/movie/{video_id}', video_id)
+        webpage = self._download_webpage(url, video_id)
+        page_store = self._extract_pagestore(webpage, video_id)
+        _, detail, metadata = self._parse_openrec_metadata(page_store, video_id)
 
-        return self._extract_movie(webpage, video_id, 'movie', False)
+        formats = []
+        for m3u8_url in traverse_obj(detail, (
+            'data', 'items', ..., 'media',
+            ('url', 'url_audio'), {url_or_none},
+        )):
+            formats.extend(self._extract_m3u8_formats(
+                m3u8_url, video_id, 'mp4', headers=self._HEADERS))
+
+        return {
+            'formats': formats,
+            **metadata,
+        }
+
+
+class OpenRecPlaylistIE(OpenRecBaseIE):
+    IE_NAME = 'mellowfan:playlist'
+
+    _VALID_URL = r'https?://(?:www\.)?(?:mellow-fan\.com|openrec\.tv)/(?:m/)?user/[^/?#]+/playlist/(?P<id>[^/?#]+)'
+    _TESTS = [{
+        # live
+        'url': 'https://www.mellow-fan.com/user/DbD_BPF/playlist/j59svruhtua2z8t',
+        'info_dict': {
+            'id': 'j59svruhtua2z8t',
+            'title': 'BPFのおすすめ',
+        },
+        'playlist_mincount': 10,
+    }, {
+        # capture
+        'url': 'https://www.mellow-fan.com/user/sagara_mayu/playlist/xngNMzv71yLjGdW',
+        'info_dict': {
+            'id': 'xngNMzv71yLjGdW',
+            'title': '相良茉優のFAN!FUN!FACTORY!キャプチャ',
+        },
+        'playlist_mincount': 4,
+    }, {
+        # movie
+        'url': 'https://www.mellow-fan.com/user/oreranohonoka/playlist/sficoshvi9dgkqh',
+        'info_dict': {
+            'id': 'sficoshvi9dgkqh',
+            'title': 'SP（スペシャル）映像：（短編映像）おにいたむプランだけが視聴できます',
+        },
+        'playlist_mincount': 111,
+    }]
+
+    def _entries(self, items):
+        for movie in traverse_obj(items, (
+            'playlist_movies', ..., 'movie', 'id', {str_or_none},
+        )):
+            is_live = traverse_obj(movie, ('movie', 'is_live', {bool}))
+            path, ie = ('live', OpenRecIE) if is_live else ('movie', OpenRecMovieIE)
+            movie_id = movie['movie']['id']
+
+            yield self.url_result(f'{self._BASE_URL}/{path}/{movie_id}', ie)
+
+        for capture in traverse_obj(items, (
+            'playlist_captures', ...,
+            'capture_relation', 'capture', 'id', {str_or_none},
+        )):
+            capture_id = capture['capture_relation']['capture']['id']
+
+            yield self.url_result(
+                f'{self._BASE_URL}/capture/{capture_id}', OpenRecCaptureIE)
+
+    def _real_extract(self, url):
+        playlist_id = self._match_id(url)
+        me = self._call_api(f'users/me/playlists/{playlist_id}', playlist_id)
+
+        if items := traverse_obj(me, (
+            'data', 'items', ..., {dict}, any,
+        )):
+            return self.playlist_result(
+                self._entries(items), playlist_id,
+                traverse_obj(items, ('title', {clean_html}, filter)))
+
+        webpage = self._download_webpage(url, playlist_id)
+        playlist_title = traverse_obj(webpage, (
+            {find_element(cls='sc-1ak77bz-10')}, {clean_html}, filter))
+
+        return self.playlist_from_matches(traverse_obj(webpage, (
+            {find_elements(cls='sc-vc0xhn-0')}, ...,
+            {find_element(cls='sc-1ddd11y-0', html=True)},
+            {extract_attributes}, 'href', {str},
+        )), playlist_id, playlist_title, getter=urljoin(f'{self._BASE_URL}/'))
+
+
+class OpenRecChannelIE(OpenRecBaseIE):
+    IE_NAME = 'mellowfan:channel'
+
+    _PAGE_SIZE = 40
+    _VALID_URL = r'https?://(?:www\.)?(?:mellow-fan\.com|openrec\.tv)/(?:m/)?user/(?P<id>[^/?#]+)$'
+    _TESTS = [{
+        'url': 'https://www.mellow-fan.com/user/OPENRECPARK',
+        'info_dict': {
+            'id': 'OPENRECPARK',
+            'title': 'OPENREC PARK',
+        },
+        'playlist_mincount': 40,
+    }]
+
+    def _fetch_page(self, channel_id, page):
+        page += 1
+        search_movies = self._download_json(
+            f'{self._PUBLIC_API_BASE}/search-movies', channel_id,
+            f'Downloading page {page}', query={
+                'channel_ids': channel_id,
+                'include_live': 'true',
+                'include_upload': 'true',
+                'onair_status': '2',
+                'include_deleted': 'true',
+                'sort': 'published_at',
+                'page': str(page),
+            })
+
+        for movie in traverse_obj(search_movies, (
+            lambda _, v: str_or_none(v['movie_type']) and str_or_none(v['id']),
+        )):
+            path, ie = ('live', OpenRecIE) if movie['movie_type'] == '1' else ('movie', OpenRecMovieIE)
+
+            yield self.url_result(f'{self._BASE_URL}/{path}/{movie["id"]}', ie)
+
+    def _real_extract(self, url):
+        channel_id = self._match_id(url)
+        webpage = self._download_webpage(url, channel_id)
+        page_store = self._extract_pagestore(webpage, channel_id)
+
+        channel = traverse_obj(page_store, ('state', '_channel', {dict}))
+        movie_count = traverse_obj(channel, ('movieCount', {int_or_none}))
+
+        return self.playlist_result(InAdvancePagedList(
+            functools.partial(self._fetch_page, channel_id),
+            math.ceil(movie_count / self._PAGE_SIZE), self._PAGE_SIZE,
+        ), channel_id, traverse_obj(channel, ('user', 'name', {clean_html}, filter)))
+
+
+class OpenRecChannelSearchIE(OpenRecBaseIE):
+    IE_NAME = 'mellowfan:channel:search'
+
+    _VALID_URL = r'https?://(?:www\.)?(?:mellow-fan\.com|openrec\.tv)/(?:m/)?user/(?P<id>[^/?#]+)/search(?:/(?P<type>capture|movie))?(?:[/?#]|$)'
+    _TESTS = [{
+        'url': 'https://www.mellow-fan.com/user/indegnasen/search?search_query=%E3%82%B9%E3%82%A4%E3%82%AB',
+        'info_dict': {
+            'id': 'indegnasen',
+            'title': 'indegnasen:スイカ',
+        },
+        'playlist_count': 2,
+    }, {
+        'url': 'https://www.mellow-fan.com/user/ofurekodesu/search/movie?search_query=%E3%82%A2%E3%83%AA%E3%82%AA%E5%85%AB%E5%B0%BE',
+        'info_dict': {
+            'id': 'ofurekodesu',
+            'title': 'ofurekodesu:アリオ八尾:movie',
+        },
+        'playlist_mincount': 31,
+    }, {
+        'url': 'https://www.mellow-fan.com/user/DbD_BPF/search/capture?search_query=%E3%81%82%E3%81%A3%E3%81%95%E3%82%8A%E3%81%97%E3%82%87%E3%81%93',
+        'info_dict': {
+            'id': 'DbD_BPF',
+            'title': 'DbD_BPF:あっさりしょこ:capture',
+        },
+        'playlist_mincount': 10,
+    }]
+
+    def _entries(self, channel_id, search_type, search_query):
+        api_url = f'{self._PUBLIC_API_BASE}/search-{search_type}s'
+        type_map = {
+            'capture': OpenRecCaptureIE,
+            'live': OpenRecIE,
+            'movie': OpenRecMovieIE,
+        }
+
+        for page in itertools.count(1):
+            search_items = self._download_json(
+                api_url, channel_id, f'Downloading page {page}', query={
+                    'channel_ids': channel_id,
+                    'page': page,
+                    'search_query': search_query,
+                })
+            if not search_items:
+                break
+
+            for item in search_items:
+                item_type = 'live' if search_type == 'movie' and traverse_obj(item, ('is_live', {bool})) else search_type
+                item_id = traverse_obj(item, ((None, 'capture'), 'id', {str_or_none}, any))
+
+                yield self.url_result(
+                    f'{self._BASE_URL}/{item_type}/{item_id}', type_map[item_type])
+
+    def _real_extract(self, url):
+        channel_id, search_type = self._match_valid_url(url).group('id', 'type')
+        search_query = traverse_obj(url, (
+            {parse_qs}, 'search_query', -1, {str}, filter,
+            {require('search query', expected=True)}))
+
+        if not search_type:
+            entries = []
+            for search_type in ('capture', 'movie'):
+                search_url = update_url_query(
+                    f'{self._BASE_URL}/user/{channel_id}/search/{search_type}', {'search_query': search_query})
+                entries.append(self.url_result(search_url, OpenRecChannelSearchIE))
+
+            return self.playlist_result(
+                entries, channel_id, join_nonempty(channel_id, search_query, delim=':'))
+
+        return self.playlist_result(
+            self._entries(channel_id, search_type, search_query),
+            channel_id, join_nonempty(channel_id, search_query, search_type, delim=':'))

yt_dlp/extractor/streaks.py

@@ -5,12 +5,13 @@ from .common import InfoExtractor
 from ..networking.exceptions import HTTPError
 from ..utils import (
     ExtractorError,
+    clean_html,
     filter_dict,
     float_or_none,
+    int_or_none,
     join_nonempty,
     mimetype2ext,
     parse_iso8601,
-    unsmuggle_url,
     update_url_query,
     url_or_none,
 )
@@ -22,22 +23,28 @@ class StreaksBaseIE(InfoExtractor):
     _GEO_BYPASS = False
     _GEO_COUNTRIES = ['JP']
 
-    def _extract_from_streaks_api(self, project_id, media_id, headers=None, query=None, ssai=False, live_from_start=False):
+    def _streaks_playback_api_url(self, project_id, media_id):
+        return self._API_URL_TEMPLATE.format('playback', project_id, media_id, '')
+
+    def _download_streaks_playback_json(self, project_id, media_id, headers=None):
+        return self._download_json(
+            self._streaks_playback_api_url(project_id, media_id),
+            media_id, 'Downloading STREAKS playback API JSON', headers={
+                'Accept': 'application/json',
+                'Origin': 'https://players.streaks.jp',
+                **self.geo_verification_headers(),
+                **(headers or {}),
+            })
+
+    def _extract_from_streaks_api(self, project_id, media_id, headers=None, query=None, live_from_start=False):
         try:
-            response = self._download_json(
+            response = self._download_streaks_playback_json(project_id, media_id, headers=headers)
-                self._API_URL_TEMPLATE.format('playback', project_id, media_id, ''),
-                media_id, 'Downloading STREAKS playback API JSON', headers={
-                    'Accept': 'application/json',
-                    'Origin': 'https://players.streaks.jp',
-                    **self.geo_verification_headers(),
-                    **(headers or {}),
-                })
         except ExtractorError as e:
             if isinstance(e.cause, HTTPError) and e.cause.status in (403, 404):
                 error = self._parse_json(e.cause.response.read().decode(), media_id, fatal=False)
-                message = traverse_obj(error, ('message', {str}))
+                message = traverse_obj(error, ('message', {clean_html}, filter))
-                code = traverse_obj(error, ('code', {str}))
+                code = traverse_obj(error, ('code', {clean_html}, filter))
-                error_id = traverse_obj(error, ('id', {int}))
+                error_id = traverse_obj(error, ('id', {int_or_none}))
                 if code == 'REQUEST_FAILED':
                     if error_id == 124:
                         self.raise_geo_restricted(countries=self._GEO_COUNTRIES)
@@ -59,8 +66,12 @@ class StreaksBaseIE(InfoExtractor):
 
         formats, subtitles = [], {}
         drm_formats = False
+        sources = response['sources']
+        ssai = traverse_obj(sources, (..., 'ssai', {dict}, any))
 
-        for source in traverse_obj(response, ('sources', lambda _, v: v['src'])):
+        for source in traverse_obj(sources, (
+            lambda _, v: url_or_none(v['src']),
+        )):
             if source.get('key_systems'):
                 drm_formats = True
                 continue
@@ -108,12 +119,12 @@ class StreaksBaseIE(InfoExtractor):
             'subtitles': subtitles,
             'uploader_id': project_id,
             **traverse_obj(response, {
-                'title': ('name', {str}),
+                'title': ('name', {clean_html}, filter),
-                'description': ('description', {str}, filter),
+                'description': ('description', {clean_html}, filter),
                 'duration': ('duration', {float_or_none}),
                 'modified_timestamp': ('updated_at', {parse_iso8601}),
-                'tags': ('tags', ..., {str}),
+                'tags': ('tags', ..., {clean_html}, filter, all, filter),
-                'thumbnails': (('poster', 'thumbnail'), 'src', {'url': {url_or_none}}),
+                'thumbnail': (('thumbnail', 'poster'), 'src', {url_or_none}, any),
                 'timestamp': ('created_at', {parse_iso8601}),
             }),
         }
@@ -121,26 +132,27 @@ class StreaksBaseIE(InfoExtractor):
 
 class StreaksIE(StreaksBaseIE):
     _VALID_URL = [
-        r'https?://players\.streaks\.jp/(?P<project_id>[\w-]+)/[\da-f]+/index\.html\?(?:[^#]+&)?m=(?P<id>(?:ref:)?[\w-]+)',
+        r'https?://players\.streaks\.jp/(?P<project_id>[\w-]+)/(?P<api_key>[\da-f]+)/index\.html\?(?:[^#]+&)?m=(?P<id>(?:ref:)?[\w-]+)',
         r'https?://playback\.api\.streaks\.jp/v1/projects/(?P<project_id>[\w-]+)/medias/(?P<id>(?:ref:)?[\w-]+)',
     ]
     _EMBED_REGEX = [rf'<iframe\s+[^>]*\bsrc\s*=\s*["\'](?P<url>{_VALID_URL[0]})']
     _TESTS = [{
-        'url': 'https://players.streaks.jp/tipness/08155cd19dc14c12bebefb69b92eafcc/index.html?m=dbdf2df35b4d483ebaeeaeb38c594647',
+        # https://online.tipness.co.jp/contents/9a064492-a62b-5d03-4668-6b40d6325b1e
+        'url': 'https://players.streaks.jp/tipness/08155cd19dc14c12bebefb69b92eafcc/index.html?m=ba2c253508914d9ea061a5f26bc58b20',
         'info_dict': {
-            'id': 'dbdf2df35b4d483ebaeeaeb38c594647',
+            'id': 'ba2c253508914d9ea061a5f26bc58b20',
             'ext': 'mp4',
-            'title': '3shunenCM_edit.mp4',
+            'title': 'tarun_suimin.mp4',
-            'display_id': 'dbdf2df35b4d483ebaeeaeb38c594647',
+            'duration': 265.344,
-            'duration': 47.533,
             'live_status': 'not_live',
-            'modified_date': '20230726',
+            'modified_date': '20230908',
-            'modified_timestamp': 1690356180,
+            'modified_timestamp': 1694146842,
-            'timestamp': 1690355996,
+            'timestamp': 1694145352,
-            'upload_date': '20230726',
+            'upload_date': '20230908',
             'uploader_id': 'tipness',
         },
     }, {
+        # https://www.ktv.jp/mycoffeetime/
         'url': 'https://players.streaks.jp/ktv-web/0298e8964c164ab384c07ef6e08c444b/index.html?m=ref:mycoffeetime_250317',
         'info_dict': {
             'id': 'dccdc079e3fd41f88b0c8435e2d453ab',
@@ -157,22 +169,24 @@ class StreaksIE(StreaksBaseIE):
             'uploader_id': 'ktv-web',
         },
     }, {
-        'url': 'https://playback.api.streaks.jp/v1/projects/ktv-web/medias/b5411938e1e5435dac71edf829dd4813',
+        # https://www.ktv.jp/news/articles/?id=28105
+        'url': 'https://playback.api.streaks.jp/v1/projects/ktv-news/medias/714171c4c53c409bb41e1572997ebfcf',
         'info_dict': {
-            'id': 'b5411938e1e5435dac71edf829dd4813',
+            'id': '714171c4c53c409bb41e1572997ebfcf',
             'ext': 'mp4',
-            'title': 'KANTELE_SYUSEi_0630',
+            'title': '28105.mp4',
-            'display_id': 'b5411938e1e5435dac71edf829dd4813',
+            'duration': 49.984,
             'live_status': 'not_live',
-            'modified_date': '20250122',
+            'modified_date': '20260702',
-            'modified_timestamp': 1737522999,
+            'modified_timestamp': 1783028559,
             'thumbnail': r're:https?://.+\.jpg',
-            'timestamp': 1735205137,
+            'timestamp': 1783028407,
-            'upload_date': '20241226',
+            'upload_date': '20260702',
-            'uploader_id': 'ktv-web',
+            'uploader_id': 'ktv-news',
         },
+        'params': {'extractor_args': {'streaks': {'api_key': ['0ff2ccfb6381401582d6ee60e3cb66a1']}}},
     }, {
-        # TVer Olympics: website already down, but api remains accessible
+        # https://tver.jp/olympic/paris2024/live/FBLMTEAM11------------SFNL000100--/
         'url': 'https://playback.api.streaks.jp/v1/projects/tver-olympic/medias/ref:sp_240806_1748_dvr',
         'info_dict': {
             'id': 'c10f7345adb648cf804d7578ab93b2e3',
@@ -187,8 +201,10 @@ class StreaksIE(StreaksBaseIE):
             'upload_date': '20240804',
             'uploader_id': 'tver-olympic',
         },
+        'params': {'extractor_args': {'streaks': {'api_key': ['e09168c4383d4b18949067022558f071']}}},
+        'skip': 'Invalid URL',
     }, {
-        # TBS FREE: 24-hour stream
+        # https://cu.tbs.co.jp/simul/simul-02
         'url': 'https://playback.api.streaks.jp/v1/projects/tbs/medias/ref:simul-02',
         'info_dict': {
             'id': 'c4e83a7b48f4409a96adacec674b4e22',
@@ -202,12 +218,14 @@ class StreaksIE(StreaksBaseIE):
             'upload_date': '20240117',
             'uploader_id': 'tbs',
         },
+        'skip': 'Invalid URL',
     }, {
         # DRM protected
         'url': 'https://players.streaks.jp/sp-jbc/a12d7ee0f40c49d6a0a2bff520639677/index.html?m=5f89c62f37ee4a68be8e6e3b1396c7d8',
         'only_matching': True,
     }]
     _WEBPAGE_TESTS = [{
+        # https://players.streaks.jp/play/719af2a1d2d544e89bcad3456eeae5d9/index.html?m=2d975178293140dc8074a7fc536a7604
         'url': 'https://event.play.jp/playnext2023/',
         'info_dict': {
             'id': '2d975178293140dc8074a7fc536a7604',
@@ -222,6 +240,7 @@ class StreaksIE(StreaksBaseIE):
             'modified_date': '20250213',
             'live_status': 'not_live',
         },
+        'params': {'nocheckcertificate': True},
     }, {
         'url': 'https://wowshop.jp/Page/special/cooking_goods/?bid=wowshop&srsltid=AfmBOor_phUNoPEE_UCPiGGSCMrJE5T2US397smvsbrSdLqUxwON0el4',
         'playlist_mincount': 2,
@@ -232,13 +251,15 @@ class StreaksIE(StreaksBaseIE):
             'age_limit': 0,
             'thumbnail': 'https://wowshop.jp/Page/special/cooking_goods/images/ogp.jpg',
         },
+        'skip': 'Invalid URL',
     }]
 
     def _real_extract(self, url):
-        url, smuggled_data = unsmuggle_url(url, {})
+        mobj = self._match_valid_url(url).groupdict()
-        project_id, media_id = self._match_valid_url(url).group('project_id', 'id')
+        project_id, media_id = mobj['project_id'], mobj['id']
+        api_key = mobj.get('api_key') or self._configuration_arg('api_key', [None])[0]
 
         return self._extract_from_streaks_api(
             project_id, media_id, headers=filter_dict({
-                'X-Streaks-Api-Key': smuggled_data.get('api_key'),
+                'X-Streaks-Api-Key': api_key,
             }))

yt_dlp/options.py

@@ -511,7 +511,7 @@ def create_parser():
     general.add_option(
         '--live-from-start',
         action='store_true', dest='live_from_start',
-        help='Download livestreams from the start. Currently experimental and only supported for YouTube, Twitch, and TVer')
+        help='Download livestreams from the start. Currently experimental and only supported for YouTube, Twitch, TVer, and mellow-fan')
     general.add_option(
         '--no-live-from-start',
         action='store_false', dest='live_from_start',