diff options
Diffstat (limited to 'youtube_dl/extractor')
| -rw-r--r-- | youtube_dl/extractor/common.py | 90 | ||||
| -rw-r--r-- | youtube_dl/extractor/youtube.py | 115 |
2 files changed, 121 insertions, 84 deletions
diff --git a/youtube_dl/extractor/common.py b/youtube_dl/extractor/common.py index cb67b976d..a64fcfccc 100644 --- a/youtube_dl/extractor/common.py +++ b/youtube_dl/extractor/common.py @@ -505,7 +505,7 @@ class InfoExtractor(object): if not self._x_forwarded_for_ip: # Geo bypass mechanism is explicitly disabled by user - if not self._downloader.params.get('geo_bypass', True): + if not self.get_param('geo_bypass', True): return if not geo_bypass_context: @@ -527,7 +527,7 @@ class InfoExtractor(object): # Explicit IP block specified by user, use it right away # regardless of whether extractor is geo bypassable or not - ip_block = self._downloader.params.get('geo_bypass_ip_block', None) + ip_block = self.get_param('geo_bypass_ip_block', None) # Otherwise use random IP block from geo bypass context but only # if extractor is known as geo bypassable @@ -538,8 +538,8 @@ class InfoExtractor(object): if ip_block: self._x_forwarded_for_ip = GeoUtils.random_ipv4(ip_block) - if self._downloader.params.get('verbose', False): - self._downloader.to_screen( + if self.get_param('verbose', False): + self.to_screen( '[debug] Using fake IP %s as X-Forwarded-For.' % self._x_forwarded_for_ip) return @@ -548,7 +548,7 @@ class InfoExtractor(object): # Explicit country code specified by user, use it right away # regardless of whether extractor is geo bypassable or not - country = self._downloader.params.get('geo_bypass_country', None) + country = self.get_param('geo_bypass_country', None) # Otherwise use random country code from geo bypass context but # only if extractor is known as geo bypassable @@ -559,8 +559,8 @@ class InfoExtractor(object): if country: self._x_forwarded_for_ip = GeoUtils.random_ipv4(country) - if self._downloader.params.get('verbose', False): - self._downloader.to_screen( + if self.get_param('verbose', False): + self.to_screen( '[debug] Using fake IP %s (%s) as X-Forwarded-For.' % (self._x_forwarded_for_ip, country.upper())) @@ -586,9 +586,9 @@ class InfoExtractor(object): raise ExtractorError('An extractor error has occurred.', cause=e) def __maybe_fake_ip_and_retry(self, countries): - if (not self._downloader.params.get('geo_bypass_country', None) + if (not self.get_param('geo_bypass_country', None) and self._GEO_BYPASS - and self._downloader.params.get('geo_bypass', True) + and self.get_param('geo_bypass', True) and not self._x_forwarded_for_ip and countries): country_code = random.choice(countries) @@ -698,7 +698,7 @@ class InfoExtractor(object): if fatal: raise ExtractorError(errmsg, sys.exc_info()[2], cause=err) else: - self._downloader.report_warning(errmsg) + self.report_warning(errmsg) return False def _download_webpage_handle(self, url_or_request, video_id, note=None, errnote=None, fatal=True, encoding=None, data=None, headers={}, query={}, expected_status=None): @@ -770,11 +770,11 @@ class InfoExtractor(object): webpage_bytes = prefix + webpage_bytes if not encoding: encoding = self._guess_encoding_from_content(content_type, webpage_bytes) - if self._downloader.params.get('dump_intermediate_pages', False): + if self.get_param('dump_intermediate_pages', False): self.to_screen('Dumping request to ' + urlh.geturl()) dump = base64.b64encode(webpage_bytes).decode('ascii') - self._downloader.to_screen(dump) - if self._downloader.params.get('write_pages', False): + self.to_screen(dump) + if self.get_param('write_pages', False): basen = '%s_%s' % (video_id, urlh.geturl()) if len(basen) > 240: h = '___' + hashlib.md5(basen.encode('utf-8')).hexdigest() @@ -976,19 +976,9 @@ class InfoExtractor(object): """Print msg to screen, prefixing it with '[ie_name]'""" self._downloader.to_screen(self.__ie_msg(msg)) - def write_debug(self, msg, only_once=False, _cache=[]): + def write_debug(self, msg, only_once=False): '''Log debug message or Print message to stderr''' - if not self.get_param('verbose', False): - return - message = '[debug] ' + self.__ie_msg(msg) - logger = self.get_param('logger') - if logger: - logger.debug(message) - else: - if only_once and hash(message) in _cache: - return - self._downloader.to_stderr(message) - _cache.append(hash(message)) + self._downloader.write_debug(self.__ie_msg(msg), only_once=only_once) # name, default=None, *args, **kwargs def get_param(self, name, *args, **kwargs): @@ -1084,7 +1074,7 @@ class InfoExtractor(object): if mobj: break - if not self._downloader.params.get('no_color') and compat_os_name != 'nt' and sys.stderr.isatty(): + if not self.get_param('no_color') and compat_os_name != 'nt' and sys.stderr.isatty(): _name = '\033[0;34m%s\033[0m' % name else: _name = name @@ -1102,7 +1092,7 @@ class InfoExtractor(object): elif fatal: raise RegexNotFoundError('Unable to extract %s' % _name) else: - self._downloader.report_warning('unable to extract %s' % _name + bug_reports_message()) + self.report_warning('unable to extract %s' % _name + bug_reports_message()) return None def _search_json(self, start_pattern, string, name, video_id, **kwargs): @@ -1172,7 +1162,7 @@ class InfoExtractor(object): username = None password = None - if self._downloader.params.get('usenetrc', False): + if self.get_param('usenetrc', False): try: netrc_machine = netrc_machine or self._NETRC_MACHINE info = netrc.netrc().authenticators(netrc_machine) @@ -1183,7 +1173,7 @@ class InfoExtractor(object): raise netrc.NetrcParseError( 'No authenticators for %s' % netrc_machine) except (AttributeError, IOError, netrc.NetrcParseError) as err: - self._downloader.report_warning( + self.report_warning( 'parsing .netrc: %s' % error_to_compat_str(err)) return username, password @@ -1220,10 +1210,10 @@ class InfoExtractor(object): """ if self._downloader is None: return None - downloader_params = self._downloader.params - if downloader_params.get('twofactor') is not None: - return downloader_params['twofactor'] + twofactor = self.get_param('twofactor') + if twofactor is not None: + return twofactor return compat_getpass('Type %s and press [Return]: ' % note) @@ -1358,7 +1348,7 @@ class InfoExtractor(object): elif fatal: raise RegexNotFoundError('Unable to extract JSON-LD') else: - self._downloader.report_warning('unable to extract JSON-LD %s' % bug_reports_message()) + self.report_warning('unable to extract JSON-LD %s' % bug_reports_message()) return {} def _json_ld(self, json_ld, video_id, fatal=True, expected_type=None): @@ -1589,7 +1579,7 @@ class InfoExtractor(object): if f.get('vcodec') == 'none': # audio only preference -= 50 - if self._downloader.params.get('prefer_free_formats'): + if self.get_param('prefer_free_formats'): ORDER = ['aac', 'mp3', 'm4a', 'webm', 'ogg', 'opus'] else: ORDER = ['webm', 'opus', 'ogg', 'mp3', 'aac', 'm4a'] @@ -1601,7 +1591,7 @@ class InfoExtractor(object): else: if f.get('acodec') == 'none': # video only preference -= 40 - if self._downloader.params.get('prefer_free_formats'): + if self.get_param('prefer_free_formats'): ORDER = ['flv', 'mp4', 'webm'] else: ORDER = ['webm', 'flv', 'mp4'] @@ -1667,7 +1657,7 @@ class InfoExtractor(object): """ Either "http:" or "https:", depending on the user's preferences """ return ( 'http:' - if self._downloader.params.get('prefer_insecure', False) + if self.get_param('prefer_insecure', False) else 'https:') def _proto_relative_url(self, url, scheme=None): @@ -3199,7 +3189,7 @@ class InfoExtractor(object): if fatal: raise ExtractorError(msg) else: - self._downloader.report_warning(msg) + self.report_warning(msg) return res def _float(self, v, name, fatal=False, **kwargs): @@ -3209,7 +3199,7 @@ class InfoExtractor(object): if fatal: raise ExtractorError(msg) else: - self._downloader.report_warning(msg) + self.report_warning(msg) return res def _set_cookie(self, domain, name, value, expire_time=None, port=None, @@ -3218,12 +3208,12 @@ class InfoExtractor(object): 0, name, value, port, port is not None, domain, True, domain.startswith('.'), path, True, secure, expire_time, discard, None, None, rest) - self._downloader.cookiejar.set_cookie(cookie) + self.cookiejar.set_cookie(cookie) def _get_cookies(self, url): """ Return a compat_cookies_SimpleCookie with the cookies for the url """ req = sanitized_Request(url) - self._downloader.cookiejar.add_cookie_header(req) + self.cookiejar.add_cookie_header(req) return compat_cookies_SimpleCookie(req.get_header('Cookie')) def _apply_first_set_cookie_header(self, url_handle, cookie): @@ -3283,8 +3273,8 @@ class InfoExtractor(object): return not any_restricted def extract_subtitles(self, *args, **kwargs): - if (self._downloader.params.get('writesubtitles', False) - or self._downloader.params.get('listsubtitles')): + if (self.get_param('writesubtitles', False) + or self.get_param('listsubtitles')): return self._get_subtitles(*args, **kwargs) return {} @@ -3305,7 +3295,11 @@ class InfoExtractor(object): """ Merge subtitle dictionaries, language by language. """ # ..., * , target=None - target = kwargs.get('target') or dict(subtitle_dict1) + target = kwargs.get('target') + if target is None: + target = dict(subtitle_dict1) + else: + subtitle_dicts = (subtitle_dict1,) + subtitle_dicts for subtitle_dict in subtitle_dicts: for lang in subtitle_dict: @@ -3313,8 +3307,8 @@ class InfoExtractor(object): return target def extract_automatic_captions(self, *args, **kwargs): - if (self._downloader.params.get('writeautomaticsub', False) - or self._downloader.params.get('listsubtitles')): + if (self.get_param('writeautomaticsub', False) + or self.get_param('listsubtitles')): return self._get_automatic_captions(*args, **kwargs) return {} @@ -3322,9 +3316,9 @@ class InfoExtractor(object): raise NotImplementedError('This method must be implemented by subclasses') def mark_watched(self, *args, **kwargs): - if (self._downloader.params.get('mark_watched', False) + if (self.get_param('mark_watched', False) and (self._get_login_info()[0] is not None - or self._downloader.params.get('cookiefile') is not None)): + or self.get_param('cookiefile') is not None)): self._mark_watched(*args, **kwargs) def _mark_watched(self, *args, **kwargs): @@ -3332,7 +3326,7 @@ class InfoExtractor(object): def geo_verification_headers(self): headers = {} - geo_verification_proxy = self._downloader.params.get('geo_verification_proxy') + geo_verification_proxy = self.get_param('geo_verification_proxy') if geo_verification_proxy: headers['Ytdl-request-proxy'] = geo_verification_proxy return headers diff --git a/youtube_dl/extractor/youtube.py b/youtube_dl/extractor/youtube.py index ce97fd75b..54073ef86 100644 --- a/youtube_dl/extractor/youtube.py +++ b/youtube_dl/extractor/youtube.py @@ -342,14 +342,7 @@ class YoutubeBaseInfoExtractor(InfoExtractor): if not self._login(): return - _DEFAULT_API_DATA = { - 'context': { - 'client': { - 'clientName': 'WEB', - 'clientVersion': '2.20201021.03.00', - }, - }, - } + _DEFAULT_API_DATA = {'context': _INNERTUBE_CLIENTS['web']['INNERTUBE_CONTEXT']} _YT_INITIAL_DATA_RE = r'(?:window\s*\[\s*["\']ytInitialData["\']\s*\]|ytInitialData)\s*=\s*({.+?})\s*;' _YT_INITIAL_PLAYER_RESPONSE_RE = r'ytInitialPlayerResponse\s*=\s*({.+?})\s*;' @@ -497,11 +490,15 @@ class YoutubeBaseInfoExtractor(InfoExtractor): data['params'] = params for page_num in itertools.count(1): search = self._download_json( - 'https://www.youtube.com/youtubei/v1/search?key=AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', + 'https://www.youtube.com/youtubei/v1/search', video_id='query "%s"' % query, note='Downloading page %s' % page_num, errnote='Unable to download API page', fatal=False, data=json.dumps(data).encode('utf8'), + query={ + # 'key': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', + 'prettyPrint': 'false', + }, headers={'content-type': 'application/json'}) if not search: break @@ -1655,7 +1652,7 @@ class YoutubeIE(YoutubeBaseInfoExtractor): assert os.path.basename(func_id) == func_id self.write_debug('Extracting signature function {0}'.format(func_id)) - cache_spec, code = self.cache.load('youtube-sigfuncs', func_id), None + cache_spec, code = self.cache.load('youtube-sigfuncs', func_id, min_ver='2025.04.07'), None if not cache_spec: code = self._load_player(video_id, player_url, player_id) @@ -1816,6 +1813,10 @@ class YoutubeIE(YoutubeBaseInfoExtractor): return ret def _extract_n_function_name(self, jscode): + func_name, idx = None, None + # these special cases are redundant and probably obsolete (2025-04): + # they make the tests run ~10% faster without fallback warnings + r""" func_name, idx = self._search_regex( # (y=NuD(),Mw(k),q=k.Z[y]||null)&&(q=narray[idx](q),k.set(y,q),k.V||NuD(''))}}; # (R="nn"[+J.Z],mW(J),N=J.K[R]||null)&&(N=narray[idx](N),J.set(R,N))}}; @@ -1842,9 +1843,28 @@ class YoutubeIE(YoutubeBaseInfoExtractor): \(\s*[\w$]+\s*\) ''', jscode, 'Initial JS player n function name', group=('nfunc', 'idx'), default=(None, None)) + """ + + if not func_name: + # nfunc=function(x){...}|function nfunc(x); ... + # ... var y=[nfunc]|y[idx]=nfunc); + # obvious REs hang, so use a two-stage tactic + for m in re.finditer(r'''(?x) + [\n;]var\s(?:(?:(?!,).)+,|\s)*?(?!\d)[\w$]+(?:\[(?P<idx>\d+)\])?\s*=\s* + (?(idx)|\[\s*)(?P<nfunc>(?!\d)[\w$]+)(?(idx)|\s*\]) + \s*?[;\n] + ''', jscode): + func_name = self._search_regex( + r'[;,]\s*(function\s+)?({0})(?(1)|\s*=\s*function)\s*\((?!\d)[\w$]+\)\s*\{1}(?!\s*return\s)'.format( + re.escape(m.group('nfunc')), '{'), + jscode, 'Initial JS player n function name (2)', group=2, default=None) + if func_name: + idx = m.group('idx') + break + # thx bashonly: yt-dlp/yt-dlp/pull/10611 if not func_name: - self.report_warning('Falling back to generic n function search') + self.report_warning('Falling back to generic n function search', only_once=True) return self._search_regex( r'''(?xs) (?:(?<=[^\w$])|^) # instead of \b, which ignores $ @@ -1858,14 +1878,14 @@ class YoutubeIE(YoutubeBaseInfoExtractor): return func_name return self._search_json( - r'var\s+{0}\s*='.format(re.escape(func_name)), jscode, + r'(?<![\w-])var\s(?:(?:(?!,).)+,|\s)*?{0}\s*='.format(re.escape(func_name)), jscode, 'Initial JS player n function list ({0}.{1})'.format(func_name, idx), - func_name, contains_pattern=r'\[[\s\S]+\]', end_pattern='[,;]', + func_name, contains_pattern=r'\[.+\]', end_pattern='[,;]', transform_source=js_to_json)[int(idx)] def _extract_n_function_code(self, video_id, player_url): player_id = self._extract_player_info(player_url) - func_code = self.cache.load('youtube-nsig', player_id) + func_code = self.cache.load('youtube-nsig', player_id, min_ver='2025.04.07') jscode = func_code or self._load_player(video_id, player_url) jsi = JSInterpreter(jscode) @@ -3339,6 +3359,20 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): 'thumbnailViewModel', 'image'), final_key='sources'), }) + def _extract_shorts_lockup_view_model(self, view_model): + content_id = traverse_obj(view_model, ( + 'onTap', 'innertubeCommand', 'reelWatchEndpoint', 'videoId', + T(lambda v: v if YoutubeIE.suitable(v) else None))) + if not content_id: + return + return merge_dicts(self.url_result( + content_id, ie=YoutubeIE.ie_key(), video_id=content_id), { + 'title': traverse_obj(view_model, ( + 'overlayMetadata', 'primaryText', 'content', T(compat_str))), + 'thumbnails': self._extract_thumbnails( + view_model, 'thumbnail', final_key='sources'), + }) + def _video_entry(self, video_renderer): video_id = video_renderer.get('videoId') if video_id: @@ -3385,10 +3419,9 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): yield entry def _rich_grid_entries(self, contents): - for content in contents: - content = traverse_obj( - content, ('richItemRenderer', 'content'), - expected_type=dict) or {} + for content in traverse_obj( + contents, (Ellipsis, 'richItemRenderer', 'content'), + expected_type=dict): video_renderer = traverse_obj( content, 'videoRenderer', 'reelItemRenderer', expected_type=dict) @@ -3396,6 +3429,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): entry = self._video_entry(video_renderer) if entry: yield entry + # shorts item + shorts_lockup_view_model = content.get('shortsLockupViewModel') + if shorts_lockup_view_model: + entry = self._extract_shorts_lockup_view_model(shorts_lockup_view_model) + if entry: + yield entry # playlist renderer = traverse_obj( content, 'playlistRenderer', expected_type=dict) or {} @@ -3434,23 +3473,15 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): next_continuation = cls._extract_next_continuation_data(renderer) if next_continuation: return next_continuation - contents = [] - for key in ('contents', 'items'): - contents.extend(try_get(renderer, lambda x: x[key], list) or []) - for content in contents: - if not isinstance(content, dict): - continue - continuation_ep = try_get( - content, lambda x: x['continuationItemRenderer']['continuationEndpoint'], - dict) - if not continuation_ep: - continue - continuation = try_get( - continuation_ep, lambda x: x['continuationCommand']['token'], compat_str) + for command in traverse_obj(renderer, ( + ('contents', 'items', 'rows'), Ellipsis, 'continuationItemRenderer', + ('continuationEndpoint', ('button', 'buttonRenderer', 'command')), + (('commandExecutorCommand', 'commands', Ellipsis), None), T(dict))): + continuation = traverse_obj(command, ('continuationCommand', 'token', T(compat_str))) if not continuation: continue - ctp = continuation_ep.get('clickTrackingParams') - return YoutubeTabIE._build_continuation_query(continuation, ctp) + ctp = command.get('clickTrackingParams') + return cls._build_continuation_query(continuation, ctp) def _entries(self, tab, item_id, webpage): tab_content = try_get(tab, lambda x: x['content'], dict) @@ -3499,6 +3530,13 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): entry = self._video_entry(renderer) if entry: yield entry + renderer = isr_content.get('richGridRenderer') + if renderer: + for from_ in self._rich_grid_entries( + traverse_obj(renderer, ('contents', Ellipsis, T(dict)))): + yield from_ + continuation = self._extract_continuation(renderer) + continue if not continuation: continuation = self._extract_continuation(is_renderer) @@ -3508,8 +3546,9 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): rich_grid_renderer = tab_content.get('richGridRenderer') if not rich_grid_renderer: return - for entry in self._rich_grid_entries(rich_grid_renderer.get('contents') or []): - yield entry + for from_ in self._rich_grid_entries( + traverse_obj(rich_grid_renderer, ('contents', Ellipsis, T(dict)))): + yield from_ continuation = self._extract_continuation(rich_grid_renderer) @@ -3555,8 +3594,12 @@ class YoutubeTabIE(YoutubeBaseInfoExtractor): # Downloading page may result in intermittent 5xx HTTP error # that is usually worked around with a retry response = self._download_json( - 'https://www.youtube.com/youtubei/v1/browse?key=AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', + 'https://www.youtube.com/youtubei/v1/browse', None, 'Downloading page %d%s' % (page_num, ' (retry #%d)' % count if count else ''), + query={ + # 'key': 'AIzaSyAO_FJ2SlqU8Q4STEHLGCilw_Y9_11qcW8', + 'prettyPrint': 'false', + }, headers=headers, data=json.dumps(data).encode('utf8')) break except ExtractorError as e: |