From 21f40e75dfc0055ea9cdbd7fe2c46c6f9b561afd Mon Sep 17 00:00:00 2001 From: c-basalt <117849907+c-basalt@users.noreply.github.com> Date: Thu, 21 Sep 2023 13:34:35 -0400 Subject: [PATCH] [ie/douyutv] Fix extractors (#7652) Closes #2494, Closes #7295 Authored by: c-basalt --- yt_dlp/extractor/douyutv.py | 273 ++++++++++++++++++++++++------------ 1 file changed, 184 insertions(+), 89 deletions(-) diff --git a/yt_dlp/extractor/douyutv.py b/yt_dlp/extractor/douyutv.py index fa40844df..ee8893d5a 100644 --- a/yt_dlp/extractor/douyutv.py +++ b/yt_dlp/extractor/douyutv.py @@ -1,31 +1,72 @@ import time import hashlib -import re import urllib +import uuid from .common import InfoExtractor +from .openload import PhantomJSwrapper from ..utils import ( ExtractorError, + UserNotLive, + determine_ext, + int_or_none, + js_to_json, + parse_resolution, + str_or_none, + traverse_obj, unescapeHTML, - unified_strdate, + url_or_none, + urlencode_postdata, urljoin, ) -class DouyuTVIE(InfoExtractor): - IE_DESC = '斗鱼' +class DouyuBaseIE(InfoExtractor): + def _download_cryptojs_md5(self, video_id): + for url in [ + 'https://cdnjs.cloudflare.com/ajax/libs/crypto-js/3.1.2/rollups/md5.js', + 'https://cdn.bootcdn.net/ajax/libs/crypto-js/3.1.2/rollups/md5.js', + ]: + js_code = self._download_webpage( + url, video_id, note='Downloading signing dependency', fatal=False) + if js_code: + self.cache.store('douyu', 'crypto-js-md5', js_code) + return js_code + raise ExtractorError('Unable to download JS dependency (crypto-js/md5)') + + def _get_cryptojs_md5(self, video_id): + return self.cache.load('douyu', 'crypto-js-md5') or self._download_cryptojs_md5(video_id) + + def _calc_sign(self, sign_func, video_id, a): + b = uuid.uuid4().hex + c = round(time.time()) + js_script = f'{self._get_cryptojs_md5(video_id)};{sign_func};console.log(ub98484234("{a}","{b}","{c}"))' + phantom = PhantomJSwrapper(self) + result = phantom.execute(js_script, video_id, + note='Executing JS signing script').strip() + return {i: v[0] for i, v in urllib.parse.parse_qs(result).items()} + + def _search_js_sign_func(self, webpage, fatal=True): + # The greedy look-behind ensures last possible script tag is matched + return self._search_regex( + r'(?:]*>(.*?ub98484234.*?)', webpage, 'JS sign func', fatal=fatal) + + +class DouyuTVIE(DouyuBaseIE): + IE_DESC = '斗鱼直播' _VALID_URL = r'https?://(?:www\.)?douyu(?:tv)?\.com/(topic/\w+\?rid=|(?:[^/]+/))*(?P[A-Za-z0-9]+)' _TESTS = [{ - 'url': 'http://www.douyutv.com/iseven', + 'url': 'https://www.douyu.com/pigff', 'info_dict': { - 'id': '17732', - 'display_id': 'iseven', - 'ext': 'flv', - 'title': 're:^清晨醒脑!根本停不下来! [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$', - 'description': r're:.*m7show@163\.com.*', - 'thumbnail': r're:^https?://.*\.png', - 'uploader': '7师傅', + 'id': '24422', + 'display_id': 'pigff', + 'ext': 'mp4', + 'title': 're:^【PIGFF】.* [0-9]{4}-[0-9]{2}-[0-9]{2} [0-9]{2}:[0-9]{2}$', + 'description': r'≥15级牌子看鱼吧置顶帖进粉丝vx群', + 'thumbnail': str, + 'uploader': 'pigff', 'is_live': True, + 'live_status': 'is_live', }, 'params': { 'skip_download': True, @@ -85,15 +126,43 @@ class DouyuTVIE(InfoExtractor): 'only_matching': True, }] + def _get_sign_func(self, room_id, video_id): + return self._download_json( + f'https://www.douyu.com/swf_api/homeH5Enc?rids={room_id}', video_id, + note='Getting signing script')['data'][f'room{room_id}'] + + def _extract_stream_formats(self, stream_formats): + formats = [] + for stream_info in traverse_obj(stream_formats, (..., 'data')): + stream_url = urljoin( + traverse_obj(stream_info, 'rtmp_url'), traverse_obj(stream_info, 'rtmp_live')) + if stream_url: + rate_id = traverse_obj(stream_info, ('rate', {int_or_none})) + rate_info = traverse_obj(stream_info, ('multirates', lambda _, v: v['rate'] == rate_id), get_all=False) + ext = determine_ext(stream_url) + formats.append({ + 'url': stream_url, + 'format_id': str_or_none(rate_id), + 'ext': 'mp4' if ext == 'm3u8' else ext, + 'protocol': 'm3u8_native' if ext == 'm3u8' else 'https', + 'quality': rate_id % -10000 if rate_id is not None else None, + **traverse_obj(rate_info, { + 'format': ('name', {str_or_none}), + 'tbr': ('bit', {int_or_none}), + }), + }) + return formats + def _real_extract(self, url): video_id = self._match_id(url) - if video_id.isdigit(): - room_id = video_id - else: - page = self._download_webpage(url, video_id) - room_id = self._html_search_regex( - r'"room_id\\?"\s*:\s*(\d+),', page, 'room id') + webpage = self._download_webpage(url, video_id) + room_id = self._search_regex(r'\$ROOM\.room_id\s*=\s*(\d+)', webpage, 'room id') + + if self._search_regex(r'"videoLoop"\s*:\s*(\d+)', webpage, 'loop', default='') == '1': + raise UserNotLive('The channel is auto-playing VODs', video_id=video_id) + if self._search_regex(r'\$ROOM\.show_status\s*=\s*(\d+)', webpage, 'status', default='') == '2': + raise UserNotLive(video_id=video_id) # Grab metadata from API params = { @@ -102,110 +171,136 @@ def _real_extract(self, url): 'time': int(time.time()), } params['auth'] = hashlib.md5( - f'room/{video_id}?{urllib.parse.urlencode(params)}zNzMV1y4EMxOHS6I5WKm'.encode()).hexdigest() - room = self._download_json( + f'room/{room_id}?{urllib.parse.urlencode(params)}zNzMV1y4EMxOHS6I5WKm'.encode()).hexdigest() + room = traverse_obj(self._download_json( f'http://www.douyutv.com/api/v1/room/{room_id}', video_id, - note='Downloading room info', query=params)['data'] + note='Downloading room info', query=params, fatal=False), 'data') # 1 = live, 2 = offline - if room.get('show_status') == '2': - raise ExtractorError('Live stream is offline', expected=True) + if traverse_obj(room, 'show_status') == '2': + raise UserNotLive(video_id=video_id) - video_url = urljoin('https://hls3-akm.douyucdn.cn/', self._search_regex(r'(live/.*)', room['hls_url'], 'URL')) - formats, subs = self._extract_m3u8_formats_and_subtitles(video_url, room_id) + js_sign_func = self._search_js_sign_func(webpage, fatal=False) or self._get_sign_func(room_id, video_id) + form_data = { + 'rate': 0, + **self._calc_sign(js_sign_func, video_id, room_id), + } + stream_formats = [self._download_json( + f'https://www.douyu.com/lapi/live/getH5Play/{room_id}', + video_id, note="Downloading livestream format", + data=urlencode_postdata(form_data))] - title = unescapeHTML(room['room_name']) - description = room.get('show_details') - thumbnail = room.get('room_src') - uploader = room.get('nickname') + for rate_id in traverse_obj(stream_formats[0], ('data', 'multirates', ..., 'rate')): + if rate_id != traverse_obj(stream_formats[0], ('data', 'rate')): + form_data['rate'] = rate_id + stream_formats.append(self._download_json( + f'https://www.douyu.com/lapi/live/getH5Play/{room_id}', + video_id, note=f'Downloading livestream format {rate_id}', + data=urlencode_postdata(form_data))) return { 'id': room_id, - 'display_id': video_id, - 'title': title, - 'description': description, - 'thumbnail': thumbnail, - 'uploader': uploader, + 'formats': self._extract_stream_formats(stream_formats), 'is_live': True, - 'subtitles': subs, - 'formats': formats, + **traverse_obj(room, { + 'display_id': ('url', {str}, {lambda i: i[1:]}), + 'title': ('room_name', {unescapeHTML}), + 'description': ('show_details', {str}), + 'uploader': ('nickname', {str}), + 'thumbnail': ('room_src', {url_or_none}), + }) } -class DouyuShowIE(InfoExtractor): +class DouyuShowIE(DouyuBaseIE): _VALID_URL = r'https?://v(?:mobile)?\.douyu\.com/show/(?P[0-9a-zA-Z]+)' _TESTS = [{ - 'url': 'https://v.douyu.com/show/rjNBdvnVXNzvE2yw', - 'md5': '0c2cfd068ee2afe657801269b2d86214', + 'url': 'https://v.douyu.com/show/mPyq7oVNe5Yv1gLY', 'info_dict': { - 'id': 'rjNBdvnVXNzvE2yw', + 'id': 'mPyq7oVNe5Yv1gLY', 'ext': 'mp4', - 'title': '陈一发儿:砒霜 我有个室友系列!04-01 22点场', - 'duration': 7150.08, - 'thumbnail': r're:^https?://.*\.jpg$', - 'uploader': '陈一发儿', - 'uploader_id': 'XrZwYelr5wbK', - 'uploader_url': 'https://v.douyu.com/author/XrZwYelr5wbK', - 'upload_date': '20170402', + 'title': '四川人小时候的味道“蒜苗回锅肉”,传统菜不能丢,要常做来吃', + 'duration': 633, + 'thumbnail': str, + 'uploader': '美食作家王刚V', + 'uploader_id': 'OVAO4NVx1m7Q', + 'timestamp': 1661850002, + 'upload_date': '20220830', + 'view_count': int, + 'tags': ['美食', '美食综合'], }, }, { 'url': 'https://vmobile.douyu.com/show/rjNBdvnVXNzvE2yw', 'only_matching': True, }] + _FORMATS = { + 'super': '原画', + 'high': '超清', + 'normal': '高清', + } + + _QUALITIES = { + 'super': -1, + 'high': -2, + 'normal': -3, + } + + _RESOLUTIONS = { + 'super': '1920x1080', + 'high': '1280x720', + 'normal': '852x480', + } + def _real_extract(self, url): url = url.replace('vmobile.', 'v.') video_id = self._match_id(url) webpage = self._download_webpage(url, video_id) - room_info = self._parse_json(self._search_regex( - r'var\s+\$ROOM\s*=\s*({.+});', webpage, 'room info'), video_id) + video_info = self._search_json( + r'