gE6dZddlZddlZddlZddlmZddlmZddlm Z m Z m Z m Z m Z ddlmZmZmZmZddlmZddlmZmZmZdd lmZdd lmZdd lmZmZej>e Z!d e"fd Z#dZ$dZ%d e"de&fdZ'd e"de"e"ffdZ(de"de"fdZ)de"de"fdZ*de"de"fdZ+de"de"de"fdZ,de"de"de"fdZ-dede"fdZ.de"de"fdZ/d e"de e"e e"ffd!Z0de"de fd"Z1de"de fd#Z2de"de"fd$Z3d%e d&e d'e"ddfd(Z4d)e ddfd*Z5d e"de"fd+Z6d e"de"fd,Z7de efd-Z8y).zBThis module contains all non-cipher related data extraction logic.N) OrderedDict)datetime)AnyDictListOptionalTuple)parse_qsquote urlencodeurlparse)Cipher)HTMLParseErrorLiveStreamErrorRegexMatchError regex_search)YouTubeMetadata)parse_for_objectparse_for_all_objects watch_htmlcj td|d}tj|dS#t$rYywxYw)zExtract publish date :param str watch_html: The html contents of the watch page. :rtype: str :returns: Publish date of the video. z;(?<=itemprop=\"datePublished\" content=\")\d{4}-\d{2}-\d{2}rgroupNz%Y-%m-%d)rrrstrptime)rresults W/var/www/it7/html/youtubeDownloader/venv/lib/python3.12/site-packages/pytube/extract.py publish_datersB J a    VZ 00 s & 22c$dg}|D]}||vsyy)zCheck if live stream recording is available. :param str watch_html: The html contents of the watch page. :rtype: bool :returns: Whether or not the content is private. z,This live stream recording is not available.FT)runavailable_stringsstrings rrecording_availabler#&s0 7& Z  c&gd}|D]}||vsyy)zCheck if content is private. :param str watch_html: The html contents of the watch page. :rtype: bool :returns: Whether or not the content is private. )zFThis is a private video. Please sign in to verify that you may see it.z"simpleText":"Private video"zThis video is private.TFr )rprivate_stringsr"s r is_privater'8s*O " Z  r$returnc@ td|dy#t$rYywxYw)zCheck if content is age restricted. :param str watch_html: The html contents of the watch page. :rtype: bool :returns: Whether or not the content is age restricted. zog:restrictions:agerrFT)rr)rs ris_age_restrictedr*Ls-+ZqA  s  ct|}|jdi}d|vryd|vrd|vr |d|dgfSd|vr |d|dfSddgfS)aReturn the playability status and status explanation of a video. For example, a video may have a status of LOGIN_REQUIRED, and an explanation of "This is a private video. Please sign in to verify that you may see it." This explanation is what gets incorporated into the media player overlay. :param str watch_html: The html contents of the watch page. :rtype: bool :returns: Playability status and reason of the video. playabilityStatusliveStreamability) LIVE_STREAMzVideo is a live stream.statusreasonmessagesN)initial_player_responseget)rplayer_response status_dicts rplayability_statusr6\s}.j9O!%%&92>Kk)7; { "x(;x+@*AA A  $x(+j*AA A $<r$urlctd|dS)arExtract the ``video_id`` from a YouTube url. This function supports the following patterns: - :samp:`https://youtube.com/watch?v={video_id}` - :samp:`https://youtube.com/embed/{video_id}` - :samp:`https://youtu.be/{video_id}` :param str url: A YouTube url containing a video id. :rtype: str :returns: YouTube video id. z(?:v=|\/)([0-9A-Za-z_-]{11}).*rr)r7s rvideo_idr:vs 93a HHr$cvtjj|}t|jddS)aoExtract the ``playlist_id`` from a YouTube url. This function supports the following patterns: - :samp:`https://youtube.com/playlist?list={playlist_id}` - :samp:`https://youtube.com/watch?v={video_id}&list={playlist_id}` :param str url: A YouTube url containing a playlist id. :rtype: str :returns: YouTube playlist id. listr)urllibparser r query)r7parseds r playlist_idrAs1\\ " "3 'F FLL !& )! ,,r$cgd}|D]l}tj|}|j|}|s,tj d||j d}|j d}d|d|cSt dd) aExtract the ``channel_name`` or ``channel_id`` from a YouTube url. This function supports the following patterns: - :samp:`https://youtube.com/c/{channel_name}/*` - :samp:`https://youtube.com/channel/{channel_id}/* - :samp:`https://youtube.com/u/{channel_name}/*` - :samp:`https://youtube.com/user/{channel_id}/* :param str url: A YouTube url containing a channel name. :rtype: str :returns: YouTube channel name. )z(?:\/(c)\/([%\d\w_\-]+)(\/.*)?)z%(?:\/(channel)\/([%\w\d_\-]+)(\/.*)?)z(?:\/(u)\/([%\d\w_\-]+)(\/.*)?)z"(?:\/(user)\/([%\w\d_\-]+)(\/.*)?)"finished regex search, matched: %sr9/ channel_namepatternscallerpatternrecompilesearchloggerdebugrr)r7rGrJregexfunction_match uri_styleuri_identifiers rrFrFs H 4 7#c*  LL=w G&,,Q/I+11!4Nyk>"23 34 z r$r: watch_urlc Vtd|fddt|fddddg}t|S)aConstruct the video_info url. :param str video_id: A YouTube video identifier. :param str watch_url: A YouTube watch url. :rtype: str :returns: :samp:`https://youtube.com/get_video_info` with necessary GET parameters. r:)psdefaulteurl)hlen_UShtml51cTVHTML5cverz 7.20201028)rr _video_info_url)r:rUparamss rvideo_info_urlrfsD  "  U9% &    "  F 6 ""r$ embed_htmlc td|d}d|}td|fd|fd|fd d d g}t|S#t$rd}Y4wxYw) a<Construct the video_info url. :param str video_id: A YouTube video identifier. :param str embed_html: The html contents of the embed page (for age restricted videos). :rtype: str :returns: :samp:`https://youtube.com/get_video_info` with necessary GET parameters. z"sts"\s*:\s*(\d+)r9rz!https://youtube.googleapis.com/v/r:rYstsr\r_rb)rrrrd)r:rgrjrYres rvideo_info_url_age_restrictedrksu/1E /xj 9D   " TN CL   "   F 6 "" s7 AArecdt|zS)Nz'https://www.youtube.com/get_video_info?)r )res rrdrds 4y7H HHr$htmlcx t|dd}d|zS#ttf$rt|}Yd|zSwxYw)zGet the base JavaScript url. Construct the base JavaScript url, which contains the decipher "transforms". :param str html: The html contents of the watch page. assetsjszhttps://youtube.com)get_ytplayer_configKeyErrorrget_ytplayer_js)rmbase_jss rjs_urlrusS(%d+H5d; !7 ** o &(!$' 7 **(s 99mime_type_codeccd}tj|}|j|}|s td||j \}}||j dDcgc]}|j c}fScc}w)aParse the type data. Breaks up the data in the ``type`` key of the manifest, which contains the mime type and codecs serialized together, and splits them into separate elements. **Example**: mime_type_codec('audio/webm; codecs="opus"') -> ('audio/webm', ['opus']) :param str mime_type_codec: String containing mime type and codecs. :rtype: tuple :returns: The mime type and a list of codecs. z,(\w+\/\w+)\;\scodecs=\"([a-zA-Z-0-9.,\s]*)\"rvrH,)rLrMrNrgroupssplitstrip)rvrJrQresults mime_typecodecsr`s rrvrv so$>G JJw Ell?+G %6HH(Iv &,,s*;X_ LL9 : ((F3K(H  96&> c]   3 34 +113 !AqtG  ( U |002 2\#./I&&y1E %L ""#3z'8'8&9*//9J!IVbLcKde%(5!W+( 1 0"6() %i00  14 sE5F7E=<E= stream_datacjd|vryg}d|jvr|j|dd|jvr|j|d|D]E}d|vr(d|vr$t|d}|dd|d<|dd|d<|jdd k(|d <Gtj d |S) a-Apply various in-place transforms to YouTube's media stream data. Creates a ``list`` of dictionaries by string splitting on commas, then taking each list item, parsing it as a query string, converting it to a ``dict`` and unquoting the value. :param dict stream_data: Dictionary containing query string encoded values. **Example**: >>> d = {'foo': 'bar=1&var=test,em=5&t=url%20encoded'} >>> apply_descrambler(d, 'foo') >>> print(d) {'foo': [{'bar': '1', 'var': 'test'}, {'em': '5', 't': 'url encoded'}]} r7NformatsadaptiveFormatssignatureCipherrrtypeFORMAT_STREAM_TYPE_OTFis_otfzapplying descrambler)rextendr r3rOrP)rrdata cipher_urls rapply_descramblerrs$ GK$$&&{9-.K,,..{#456F   D(%d+<&=> (/2U &sOA.S &)-EEX F LL'( Nr$clddg}|D]} t||cStdd#t$rY+wxYw)zExtract the ytInitialData json from the watch_html page. This mostly contains metadata necessary for rendering the page on-load, such as video information, copyright notices, etc. @param watch_html: Html of the watch page @return: z'window\[['\"]ytInitialData['\"]]\s*=\s*zytInitialData\s*=\s* initial_datainitial_data_patternrHrrrrrGrJs rrrsX 3H #J8 8 9O PP    ' 33clddg}|D]} t||cStdd#t$rY+wxYw)aExtract the ytInitialPlayerResponse json from the watch_html page. This mostly contains metadata necessary for rendering the page on-load, such as video information, copyright notices, etc. @param watch_html: Html of the watch page @return: z1window\[['\"]ytInitialPlayerResponse['\"]]\s*=\s*rr2initial_player_response_patternrHrrs rr2r2s] =)H #J8 8 (1    rc |dddddddddd}td |}|Dcgc]}|d  }}t|S#ttf$rtgcYSwxYwcc}w) u<Get the informational metadata for the video. e.g.: [ { 'Song': '강남스타일(Gangnam Style)', 'Artist': 'PSY', 'Album': 'PSY SIX RULES Pt.1', 'Licensed to YouTube by': 'YG Entertainment Inc. [...]' } ] :rtype: YouTubeMetadata contentstwoColumnWatchNextResultsr|r9videoSecondaryInfoRenderermetadataRowContainermetadataRowContainerRendererrowsc&d|jvS)NmetadataRowRenderer)r)xs rzmetadata..;s'16683r$r)rr IndexErrorrfilter)r metadata_rowsrs rmetadatar!s#*:67RS  "",../11MO "$$BDDJL 3M8EE!Q,-EME = )) j !#r""#Fs A A, A)(A))9__doc__logging urllib.parser=rL collectionsrrtypingrrrrr r r r r pytube.cipherrpytube.exceptionsrrrpytube.helpersrpytube.metadatar pytube.parserrr getLogger__name__rOstrrr#r'boolr*r6r:rArFrfrkrdrurvrsrqrrrrr2rr r$rrsH #33== NN'+A   8 $1S1$$( # $  3C:4I#I#I$-S-S-$!c!c!H#S#S#S#4#C#S#S#@IKICI + + + =S=U3S >-B=6##2+c+c+\CCD6(T6(T6(s6(t6(r&4&D&RQSQSQ,2"*h7"*r$