tjddlmZddlZddlZddlZddlZddlmZmZddl m Z ddl m Z m Z mZmZmZmZmZddlZddlmZddlmZddlmZdd lmZmZdd lmZmZdd l m!Z!ej"e#Z$d Z%d Z&dZ'eGddZ(eGddZ)eGddZ*eGddZ+GddeZ,GddeZ-GddeZ.dS)) annotationsN) dataclassfield)md5)AnyIterableIteratorListOptionalTupleType)CallbackManagerForRetrieverRunDocument) Embeddings)RunnableRunnableConfig) VectorStoreVectorStoreRetriever) ConfigDictivAivAivAcZeZdZUdZdZded<dZded<dZd ed <d Zd ed <dZ ded <dS) SummaryConfigajConfiguration for summary generation. is_enabled: True if summary is enabled, False otherwise max_results: maximum number of results to summarize response_lang: requested language for the summary prompt_name: name of the prompt to use for summarization (see https://docs.vectara.com/docs/learn/grounded-generation/select-a-summarizer) Fbool is_enabledint max_resultsengstr response_langz"vectara-summary-ext-24-05-med-omni prompt_namestreamN) __name__ __module__ __qualname____doc__r__annotations__rr r!r"f/var/www/html/CA-Chatbot/venv/lib/python3.11/site-packages/langchain_community/vectorstores/vectara.pyrrssJKM;K;;;;Fr)rc>eZdZUdZdZded<dZded<dZd ed <d S) MMRConfigaConfiguration for Maximal Marginal Relevance (MMR) search. This will soon be deprated in favor of RerankConfig. is_enabled: True if MMR is enabled, False otherwise mmr_k: number of results to fetch for MMR, defaults to 50 diversity_bias: number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to minimum diversity and 1 to maximum diversity. Defaults to 0.3. Note: diversity_bias is equivalent 1-lambda_mult where lambda_mult is the value often used in max_marginal_relevance_search() We chose to use that since we believe it's more intuitive to the user. Frr2rmmr_k333333?floatdiversity_biasN)r#r$r%r&rr'r.r1r(r)r*r,r,.sM  JEOOOONr)r,cLeZdZUdZdZded<dZded<dZd ed <d Zded <d S) RerankConfigaConfiguration for Reranker. reranker: "mmr", "rerank_multilingual_v1", "udf" or "none" rerank_k: number of results to fetch before reranking, defaults to 50 mmr_diversity_bias: for MMR only - a number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to minimum diversity and 1 to maximum diversity. Defaults to 0.3. Note: mmr_diversity_bias is equivalent 1-lambda_mult where lambda_mult is the value often used in max_marginal_relevance_search() We chose to use that since we believe it's more intuitive to the user. user_function: for UDF only - the user function to use for reranking. nonerrerankerr-rrerank_kr/r0mmr_diversity_bias user_functionN) r#r$r%r&r5r'r6r7r9r(r)r*r3r3Csc  HH #####Mr)r3ceZdZUdZdZded<dZded<dZd ed <d Zd ed <dZ ded<dZ ded<e e Z ded<e eZded< dddZd S)VectaraQueryConfigaWConfiguration for Vectara query. k: Number of Documents to return. Defaults to 10. lambda_val: lexical match parameter for hybrid search. filter Dictionary of argument(s) to filter on metadata. For example a filter can be "doc.rating > 3.0 and part.lang = 'deu'"} see https://docs.vectara.com/docs/search-apis/sql/filter-overview for more details. score_threshold: minimal score threshold for the result. If defined, results with score less than this value will be filtered out. n_sentence_before: number of sentences before the matching segment to add, defaults to 2 n_sentence_after: number of sentences before the matching segment to add, defaults to 2 rerank_config: RerankConfig configuration dataclass summary_config: SummaryConfig configuration dataclass rkr0 lambda_valr8rfilterNOptional[float]score_thresholdn_sentence_beforen_sentence_after)default_factoryr3 rerank_configrsummary_confign_sentence_context Optional[int] mmr_configOptional[MMRConfig]Optional[SummaryConfig]Optional[RerankConfig]c ||_||_||_||_| r| |_nt |_|r)||_||_tj dtn||_||_| r | |_ dS|r=td|j |j|_ tj dtdSt|_ dS)Nz[n_sentence_context is deprecated. Please use n_sentence_before and n_sentence_after insteadmmrr5r6r7z9MMRConfig is deprecated. Please use RerankConfig instead.)r=r?r@rBrHrrDrEwarningswarnDeprecationWarningrGr3r.r1) selfr=r?r@rBrDrErIrKrHrGs r*__init__zVectaraQueryConfig.__init__ws$ .  2"0D  "///D   5%7D "$6D ! ML"     &7D "$4D !  0!.D     0!-#)#-#<"""D  MK"      ".D   r)) r<r>r8NrCrCNNNN)r=rr?r0r@rrBrArDrrErrIrJrKrLrHrMrGrN)r#r$r%r&r=r'r?r@rBrDrErr3rGrrHrVr(r)r*r;r;Ys&AKKKKJF'+O++++"'% "E"E"EMEEEE$)E-$H$H$HNHHHH+/!" !,0*.260420202020202020r)r;ceZdZdZ dFdGd ZedHdZdIdZdJdZdKdLdZ dMdNdZ dMdOd%Z dPdQd)Z dRdSd/Z dTd1ZdUd2ZdVd4Z dWdXd:Ze dPdYd>Ze dPdZd@Zd[dBZd[dCZd\dEZdS)]Vectaraa`Vectara API` vector store. See (https://vectara.com). Example: .. code-block:: python from langchain_community.vectorstores import Vectara vectorstore = Vectara( vectara_customer_id=vectara_customer_id, vectara_corpus_id=vectara_corpus_id, vectara_api_key=vectara_api_key ) Nx langchainvectara_customer_id Optional[str]vectara_corpus_idvectara_api_keyvectara_api_timeoutrsourcercP|ptjd|_|ptjd|_|ptjd|_|j|j|jt dn"t d|j||_ tj |_ tj d}|j d |||_dS) zInitialize with Vectara API.VECTARA_CUSTOMER_IDVECTARA_CORPUS_IDVECTARA_API_KEYNzHCan't find Vectara credentials, customer_id or corpus_id in environment.zUsing corpus id ) max_retrieszhttp://)osenvironget_vectara_customer_id_vectara_corpus_id_vectara_api_keyloggerwarningdebug_sourcerequestsSession_sessionadapters HTTPAdaptermountr_)rUr[r]r^r_r`adapters r*rVzVectara.__init__s%8% 2:>> !< < !#4# rz~~ 8 8 !0 T2:>>BS3T3T  % -&.$, NN     LLED,CEE F F F  (** #//A/>> Iw///#6   r)returnOptional[Embeddings]cdSNr(rUs r* embeddingszVectara.embeddingsstr)dictc.|j|jd|jdS)z=Returns headers that should be attached to each post request.zapplication/json)z x-api-keyz customer-id Content-TypezX-Source)rlrjrpr|s r*_get_post_headerszVectara._get_post_headerss'.4.    r)doc_idrc :|j|j|d}|jdt j|d||j}|jdkr7t d|d|jd|j d |j d SdS) z Delete a document from the Vectara corpus. Args: doc_id (str): ID of the document to delete. Returns: bool: True if deletion was successful, False otherwise. ) customer_id corpus_id document_idz$https://api.vectara.io/v1/delete-docT)dataverifyheaderstimeoutz#Delete request failed for doc_id = z with status code , reason z, text F) rjrkrspostjsondumpsrr_ status_codermerrorreasontext)rUrbodyresponses r* _delete_doczVectara._delete_docs 40!   =%% 2D!!**,,, &    3 & & LL#f##'##2:/##=##    5tr)Fdoc use_core_apicji}|j|d<|j|d<||d<|rdnd}|j||t j||jd}|j}|}d|vr|dd nd}|d ks|r|d krd S|r|d krdSdS)Nrrdocumentz$https://api.vectara.io/v1/core/indexzhttps://api.vectara.io/v1/indexT)rurlrrrstatuscodeALREADY_EXISTSE_ALREADY_EXISTS FORBIDDENE_NO_PERMISSIONS E_SUCCEEDED) rjrkrsrrrrr_r) rUrrrequest api_endpointrrresult status_strs r* _index_doczVectara._index_doc s"$!%!: #6 !  3 2 22  =%%**,,G$$, &  * 19V1C1CVH%f-- #   ?O1O1O%%  !Z;66%% =r)idsOptional[List[str]]kwargsrOptional[bool]c F|rfd|D}t|SdS)zDelete by vector ID or other criteria. Args: ids: List of ids to delete. Returns: Optional[bool]: True if deletion is successful, False otherwise, None if not implemented. c:g|]}|Sr()r).0idrUs r* z"Vectara.delete..2s':::t''++:::r)T)all)rUrrsuccesss` r*deletezVectara.delete(s7  ::::c:::Gw<< 4r) files_list Iterable[str] metadatasOptional[List[dict]] List[str]c Zg}t|D]\}}tj|std|dD|r||ni}|t |dftj|d}| } | d|j d|j d|jd|d | |j } | jd krC| d d } td|d| d#| jdkr7| d d } || etd|d| |S)ac Vectara provides a way to add documents directly via our API where pre-processing and chunking occurs internally in an optimal way This method provides a way to use that API in LangChain Args: files_list: Iterable of strings, each representing a local file path. Files could be text, HTML, PDF, markdown, doc/docx, ppt/pptx, etc. see API docs for full list metadatas: Optional list of metadatas associated with each file Returns: List of ids associated with each of the files indexed zFile z does not exist, skippingrb)file doc_metadatarz https://api.vectara.io/upload?c=z&o=z&d=TrueT)filesrrrrr documentIdz# already exists on Vectara (doc_id=z ), skippingrzError indexing file z: ) enumeratergpathexistsrmropenrrrpoprsrrjrkr_rinfoappend) rUrrrdoc_idsinxrmdrrrrs r* add_fileszVectara.add_files7s(":.. N NIC7>>$''  DTDDDEEE#,43"BtD$//0 $ 2E,,..G KK ' ' '}))q43LqqQUQhqqq0 *H#s**!4\B XDXXVXXX%,,!4\Bv&&&& L4LL8==??LLMMMMr)textsrOptional[dict]c .t}|D])}||*|}| d|D}|rd|d<nddi}|dd}|rdnd} d |d t j|| d t||Di} || | } | d kr+| ||| n| dkrtd|gS)aRun more texts through the embeddings and add to the vectorstore. Args: texts: Iterable of strings to add to the vectorstore. metadatas: Optional list of metadatas associated with the texts. doc_metadata: optional metadata for the document This function indexes all the input text strings in the Vectara corpus as a single Vectara document, where each input text is considered a "section" and the metadata are associated with each section. if 'doc_metadata' is provided, it is associated with the Vectara document. Returns: document ID of the document added Ncg|]}iSr(r()r_s r*rz%Vectara.add_texts..s++++++r)rZr`rFpartssectionr metadataJsoncBg|]\}}|tj|dS))rr)rr)rrrs r*rz%Vectara.add_texts..s<D"tz"~~>>r))rrrziNo permissions to add document to Vectara. Check your corpus ID, customer ID and API key) rupdateencode hexdigestrirrziprrprint) rUrrrrdoc_hashtrr section_keyr success_strs r* add_textszVectara.add_textsls`.55 ( (A OOAHHJJ ' ' ' '##%%  ++U+++I  3%0L " "$k2Lzz.%88 !-.s5W: 666666r) responseSetrrrc,i|]}|d|dSnamevaluer(rms r* z)Vectara.vectara_query..%s"???A!F)QwZ???r)metadata documentIndexc,i|]}|d|dSr r(rs r*rz)Vectara.vectara_query..'s"TTTai7TTTr)r`vectaracRg|]$\}}t|d||df%Sr page_contentrrrrxrs r*rz)Vectara.vectara_query..-sU    2 !"6'      r)rPrNrrfactualConsistencyrT)rfcsrr>)rrsrrrrr_rrmrrrrBrrrrGr5r=rHrr)rUrrrrrr responses documentsrrrdoc_numdoc_mdresrrs ` r* vectara_queryzVectara.vectara_querys$t#E6<.u222Q222r))r&)rUrrdocs_and_scoress r*similarity_searchzVectara.similarity_searchcs><$;     32/2222r)r-?fetch_k lambda_multr0c Ntd|d|z |d<|j|fi|S)aSReturn docs selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to query AND diversity among selected documents. Args: query: Text to look up documents similar to. k: Number of Documents to return. Defaults to 5. fetch_k: Number of Documents to fetch to pass to MMR algorithm. Defaults to 50 lambda_mult: Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5. kwargs: any other querying variable in VectaraQueryConfig Returns: List of Documents selected by maximal marginal relevance. rPrQrG)r3r-)rUrr/r0rs r*max_marginal_relevance_searchz%Vectara.max_marginal_relevance_searchwsE0#/W[# # # &t%e66v666r)cls Type[Vectara] embeddingc d|di}|di|}|j||fd|i||S)aConstruct Vectara wrapper from raw documents. This is intended to be a quick way to get started. Example: .. code-block:: python from langchain_community.vectorstores import Vectara vectara = Vectara.from_texts( texts, vectara_customer_id=customer_id, vectara_corpus_id=corpus_id, vectara_api_key=api_key, ) rr()rr)r4rr6rrrrs r* from_textszVectara.from_textssO6zz."55 #----%PPPPPPr)rc B|di|}||||S)aConstruct Vectara wrapper from raw documents. This is intended to be a quick way to get started. Example: .. code-block:: python from langchain_community.vectorstores import Vectara vectara = Vectara.from_files( files_list, vectara_customer_id=customer_id, vectara_corpus_id=corpus_id, vectara_api_key=api_key, ) r()r)r4rr6rrrs r* from_fileszVectara.from_filess0.#----%+++r) VectaraRAGc"t||S)zReturn a Vectara RAG runnable.r;rUrs r*as_ragzVectara.as_rags$'''r)c&t||dS)z'Return a Vectara RAG runnable for chat.T)rr=r>s r*as_chatzVectara.as_chats$T2222r)VectaraRetrieverc dt||dtS)zreturn a retriever object.r) vectorstorer)rBrir;)rUrs r* as_retrieverzVectara.as_retrievers3VZZ:L:N:N%O%O    r))NNNrYrZ) r[r\r]r\r^r\r_rr`r)rxry)rxr~)rrrxrF)rr~rrrxrr{)rrrrrxr)rrrrrrrxr)NN) rrrrrrrrrxr)FN) rrrr;rrrr\rrrxr~)rrrr;rrrxr)rrrrrxr)rrrrrxr')r-r.) rrr/rr0r0rrrxr') r4r5rrr6ryrrrrrxrX) r4r5rrr6ryrrrrrxrX)rr;rxr;)rrrxrB)r#r$r%r&rVpropertyr}rrrrrrrr#r&r-r3 classmethodr8r:r?rArEr(r)r*rXrXs#$.2+/)-#&! 7 7 7 7 7DX    <!!!!!<     $+/33333p+/'+ 77777z %&* QQQQQfMMMM^83333. 77777:+/*. [>+/*. [4((((3333      r)rXcTeZdZUdZded< ded< edZddZddZdS)rBzVectara Retriever class.rXrDr;rT)arbitrary_types_allowedrr run_managerrrrrxr'c L|jj||jfi|}d|DS)Ncg|]\}}|Sr(r(r*s r*rz.r+r))rDr#r)rUrrKrr,s r*_get_relevant_documentsz(VectaraRetriever._get_relevant_documentss99$*8 VVvVV22/2222r)rrc (|jj|fi|S)zAdd documents to vectorstore.)rD add_documents)rUrrs r*rPzVectaraRetriever.add_documentss -t-iBB6BBBr)N)rrrKrrrrxr')rr'rrrxr) r#r$r%r&r'r model_configrNrPr(r)r*rBrBs""++: $L3333 CCCCCCr)rBc6eZdZdZ ddd Z dddZ dddZd S)r;zVectara RAG runnable. Parameters: vectara: Vectara object config: VectaraQueryConfig object chat: bool, default False FrrXrr;rrc>||_||_||_d|_dSr{)rrrconv_id)rUrrrs r*rVzVectaraRAG.__init__s$    r)NinputrOptional[RunnableConfig]rrrxIterator[dict]c +6Kj|jjj}jjjdtj |jj d}|j dkr6t dd|j d|jd|jd d Sg}g}d |iV|D]}|rtj|d } | d } | d} | | dd } | Yt)| ddkrJt d| ddd| dd } | rd| dd rN| d}td||dkr#d _t dE| r| dd nd }|r|_| dd r2| didd }d|iVt-| d}d|iV͉jjrfd| dD}n| d}| d }g}|D]g}d!|d"D}|d#}d$||d"D}d%|vrd&|d%<||||hd't5||D}jjjd(vr|d jj}d)|iVd S)*aGet streaming output from Vectara RAG. Args: input: The input query config: RunnableConfig object kwargs: Any additional arguments Returns: The output dictionary with question, answer and context z&https://api.vectara.io/v1/stream-queryT)rrrrr"rrrrrrNquestionzutf-8rr rrrz&Summary generation failed with status statusDetailrzChat query failed with code RESOURCE_EXHAUSTEDz-Sorry, Vectara chat turns exceeds plan limit.rrrrranswerc@g|]}|djjk|Sr)rrB)rrrUs r*rz%VectaraRAG.stream..Ys7%%% ! zDK,GGGGGGr)rrc,i|]}|d|dSr r(rs r*rz%VectaraRAG.stream..cs"KKKai7KKKr)rrc,i|]}|d|dSr r(rs r*rz%VectaraRAG.stream..es2""" !fIqz"""r)r`rcRg|]$\}}t|d||df%Srrrs r*rz%VectaraRAG.stream..msU   "Ar %-.vY)+gJ    r)rcontext)rrrrrTrsrrrrr_rrmrrr iter_linesloadsdecoderilenrrrBrrrrGr5r=)rUrUrrrrrrlinerr response_setrrst_coderTrchunkrrrr r!r"s` r*r"zVectaraRAG.streamse |++E4; 4<XX<(--L22448D!!L4 .    3 & & LL!$-$$$$=$$$    F  5!!!!''))N +N +DM +z$++g"6"677h%m4 '$jjD99G 7;;x0011A55 N&{{844Q7;;NKKNN!#;;vt44D%4 8 8%"&x. $L7$L$LMMM"&:::+/DL"LL O%BFPdhh'7>>>DG/'. {{#7>>!%kk*>CCGGQUVV$cl***  00E#U+++++{2=%%%%%1*%=%%% %1$< ,Z 8I "I& - -KKQz]KKK"#O"4""%.w%7 %C"""$611/8F8, &)))!((,,,,  &)I%>%>   C{09>"/DKM/2$c****r)r~c ddi}||D]j}d|vr |d|d<d|vr |d|d<"d|vr|dxx|dz cc<=d|vr |d|d<Mtd|k|S)Nr\r8rarYrzUnknown chunk type: )r"rmr)rUrUrrr"ris r*invokezVectaraRAG.invokes n[['' = =EE!!!&y!1Iu$$"' "3JU""H x0 %"5\E  ;E;;<<<< r)rF)rrXrr;rrr{)rUrrrVrrrxrW)rUrrrVrrrxr~)r#r$r%r&rVr"rkr(r)r*r;r;sJO,0vvvvvv,0r)r;)/ __future__rrloggingrgrR dataclassesrrhashlibrtypingrrr r r r r rq langchain_core.callbacks.managerrlangchain_core.documentsrlangchain_core.embeddingsrlangchain_core.runnablesrrlangchain_core.vectorstoresrrpydanticr getLoggerr#rmrrrrr,r3r;rXrBr;r(r)r*rxs""""""  ((((((((GGGGGGGGGGGGGGGGGG.-----000000========IIIIIIII  8 $ $'  "          (  * O0O0O0O0O0O0O0 O0do o o o o ko o o dCCCCC+CCC0[[[[[[[[[[r)