7|h1ddlZddlZddlZddlmZmZddlZddlmZddl m Z ejdZ ejdZ Gdde Zy) N)ListTuple)Document) BaseLoaderzBV\w+zav[0-9]+c beZdZdZ d deedededefdZdeefdZd ede ee ffd Z y ) BiliBiliLoaderz9 Load fetching transcripts from BiliBili videos. video_urlssessdatabili_jctbuvid3c||_d|_ ddlm}|r |r|r|j ||||_yyyy#t$r t dwxYw)a Initialize the loader with BiliBili video URLs and authentication cookies. if no authentication cookies are provided, the loader can't get transcripts and will only fetch videos info. Args: video_urls (List[str]): List of BiliBili video URLs. sessdata (str): SESSDATA cookie value for authentication. bili_jct (str): BILI_JCT cookie value for authentication. buvid3 (str): BUVI3 cookie value for authentication. Nr)videoTrequests package not found, please install it with `pip install bilibili-api-python`)r r r )r credential bilibili_apir ImportError Credential)selfr r r r rs l/var/www/html/test/engine/venv/lib/python3.12/site-packages/langchain_community/document_loaders/bilibili.py__init__zBiliBiliLoader.__init__sm$%  * V#..!HV/DO&,8  4  s 9Areturncg}|jD]4}|j|\}}t||}|j|6|S)z Load and return a list of documents containing video transcripts. Returns: List[Document]: List of Document objects transcripts and metadata. ) page_contentmetadata)r _get_bilibili_subs_and_inforappend)rresultsurl transcript video_infodocs rloadzBiliBiliLoader.load5sP?? C%)%E%Ec%J "J  ZHC NN3  rcRtj|} ddlm}m}|r,|j |j|j}n]tj|}|r8|j t|jdd|j}ntd|||j}|jd |i|jsd |fS||j|d }|jd g} | r| djd d } | j!dsd| z} t#j| } | j$dk(rft'j(| j*jdg} dj-| D cgc]} | d c} }d|dd|dd|}||fSt/j0d|d| j$d |fSt/j0d|dd |fS#t $r t dwxYwcc} w)zU Retrieve video information and transcript for a given BiliBili URL. r)syncrr)bvidrN)aidrz(Unable to find a valid video ID in URL: rcid subtitles subtitle_urlhttpzhttps:body contentz Video Title: titlez, description: descz Transcript: zFailed to fetch subtitles for z. HTTP Status Code: zNo subtitles found for video: z. Returning empty transcript.) BV_PATTERNsearchrr%rrVideogroupr AV_PATTERNint ValueErrorget_infoupdate get_subtitleget startswithrequests status_codejsonloadsr1joinwarningswarn)rrr&r%rvr(r subsub_listsub_urlresponseraw_sub_titlescraw_transcriptraw_transcript_with_meta_infos rrz*BiliBiliLoader._get_bilibili_subs_and_infoDs:  %  0  $// JA##C(CKKC AB$8T__KU #KC5!QRR!**,' 5#,'z> !1>>*U"34577;+ qkoonb9G%%f-"W,||G,H##s*!%H,<,@:~ MM05RS  :~g 4  @+QsH  H$ H!N)r)r)r)) __name__ __module__ __qualname____doc__rstrrrr"rdictrr#rrrsn I   @ d8n :s:uS$Y7G:r#r)rBrerEtypingrrr@langchain_core.documentsr)langchain_community.document_loaders.basercompiler4r8rrVr#rr\sJ -@RZZ ! RZZ $ nZnr#