3fi;ddlmZddlZddlZddlmZmZmZmZm Z m Z m Z ddl m Z ddlmZddlmZer ddlmZmZmZej,ZGdd eZy) ) annotationsN) TYPE_CHECKINGAnyDictIterableListOptionalTuple)Document) Embeddings) VectorStore)CreateDocumentRequestDocumentCollectionResponse SearchTypeceZdZdZ dfd ZeddZddZddZ d ddZ d ddZ d ddZ d dd Z d d d Z d! d"d Z d! d#d Z d! d#d Z d! d#dZ d! d"dZ d! d$dZ d! d$dZ d% d&dZ d% d&dZ d% d'dZ d% d'dZe d( d)dZd*d+dZxZS),ZepCloudVectorStoreau`Zep` vector store. It provides methods for adding texts or documents to the store, searching for similar documents, and deleting documents. Search scores are calculated using cosine similarity normalized to [0, 1]. Args: collection_name (str): The name of the collection in the Zep store. api_key (str): The API key for the Zep API. ct||s td ddlm}m}|||_|||_||_ |jy#t $r t dwxYw)NzR>RSJ  3 KKd2233WX 002J  3s/98A54A5c|jjj|j|jjj |j}|S)z= Create a new collection in the Zep backend. )rr*add_collectionrr+)r r/s r"r.z&ZepCloudVectorStore._create_collectionKsH ,,T-A-AB\\**99$:N:NO r&c ddlm}g}t|D]-\}}|j|||r||nd|r||nd/|S)Nr)r)contentmetadata document_id)r)r enumerateappend)r texts metadatas document_ids ZepDocument documentsids r"_generate_documents_to_addz.ZepCloudVectorStore._generate_documents_to_addSs\ C') e$ DAq   -6Yq\D3? QT  r&c |j|||}|jjj|j|}|S)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. document_ids: Optional list of document ids associated with the texts. kwargs: vectorstore specific parameters Returns: List of ids from adding the texts into the vectorstore. request)r?rr* add_documentsrr r8r9r:kwargsr<uuidss r" add_textszZepCloudVectorStore.add_textsfsI&33E9lS  %%33  )4  r&cK|j|||}|jjj|j|d{}|S7w)zARun more texts through the embeddings and add to the vectorstore.rAN)r?rr*rCrrDs r" aadd_textszZepCloudVectorStore.aadd_textssZ33E9lS ((11??  )@    sAA A Ac |dk(r|j|f||d|S|dk(r|j|f||d|Std|d)>Return docs most similar to query using specified search type. similaritykr4mmrsearch_type of ? not allowed. Expected search_type to be 'similarity' or 'mmr'.)similarity_searchmax_marginal_relevance_searchrr query search_typer4rNrEs r"searchzZepCloudVectorStore.searchs} , &)4))%R1xR6R R E !5455X17 !+/;; r&cK|dk(r|j|f||d|d{S|dk(r|j|f||d|d{Std|d767w)rKrLrMNrOrPrQ)asimilarity_searchamax_marginal_relevance_searchrrTs r"asearchzZepCloudVectorStore.asearchs , &000X17 E !<<<X17 !+/;; s!AA"AAAAc d|j|f||d|}|Dcgc]\}}| c}}Scc}}w)"Return docs most similar to query.rM(_similarity_search_with_relevance_scoresr rUrNr4rEresultsdoc_s r"rRz%ZepCloudVectorStore.similarity_searchsF@$??   -3 #**Q***s ,c .|j|f||d|S)z$Run similarity search with distance.rMr^)r rUrNr4rEs r"similarity_search_with_scorez0ZepCloudVectorStore.similarity_search_with_scores.=t<<   -3  r&c "|jjjd|j|||d|}|jxsgDcgc]<}t t |j|j|jxsdf>c}Scc}w)a Default similarity search with relevance scores. Modify if necessary in subclass. Return docs and relevance scores in the range [0, 1]. 0 is dissimilar, 1 is most similar. Args: query: input text k: Number of Documents to return. Defaults to 4. metadata: Optional, metadata filter **kwargs: kwargs to be passed to similarity search. Should include: score_threshold: Optional, a floating point value between 0 to 1 and filter the resulting set of retrieved docs Returns: List of Tuples of (doc, similarity_score) rtextlimitr4 page_contentr4r$) rr*rWrrar strr3r4scorer rUrNr4rErarbs r"r_zK|jjjd|j|||d|d{}|jxsgDcgc]<}t t |j|j|jxsdf>c}S7Zcc}ww)r]rgNrjrlr$) rr*rWrrar rmr3r4rnros r"(asimilarity_search_with_relevance_scoresz6 > 8>>c 0tjdgS)Unsupported in Zep Cloudz9similarity_search_by_vector is not supported in Zep Cloudwarningswarnr embeddingrNr4rEs r"similarity_search_by_vectorz/ZepCloudVectorStore.similarity_search_by_vector.s  QR r&c8KtjdgSw)rtz:asimilarity_search_by_vector is not supported in Zep Cloudrurxs r"asimilarity_search_by_vectorz0ZepCloudVectorStore.asimilarity_search_by_vector9s  RS c |jjjd|j|||d|d|}|jxsgDcgc],}t t |j|j.c}Scc}w)aReturn 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 4. fetch_k: Number of Documents to fetch to pass to MMR algorithm. Zep determines this automatically and this parameter is ignored. 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. metadata: Optional, metadata to filter the resulting set of retrieved docs Returns: List of Documents selected by maximal marginal relevance. rOrrhrir4rV mmr_lambdarjr$) rr*rWrrar rmr3r4 r rUrNfetch_k lambda_multr4rErar>s r"rSz1ZepCloudVectorStore.max_marginal_relevance_searchDs:/$,,''..  00"   __*  #aii.1:: F   s 1A>c "K|jjjd|j|||d|d|d{}|jxsgDcgc],}t t |j|j.c}S7Jcc}ww)z:Return docs selected using the maximal marginal relevance.rOrNrjr$) rr*rWrrar rmr3r4rs r"rZz2ZepCloudVectorStore.amax_marginal_relevance_searchps;**33::  00"    __*  #aii.1:: F    s!;BBB1B B Bc 0tjdgS)rtzEmax_marginal_relevance_search_by_vector is not supported in Zep Cloudrur ryrNrrr4rEs r"'max_marginal_relevance_search_by_vectorz;ZepCloudVectorStore.max_marginal_relevance_search_by_vectors  S  r&c8KtjdgSw)rtzFamax_marginal_relevance_search_by_vector is not supported in Zep Cloudrurs r"(amax_marginal_relevance_search_by_vectorz-1 + +  ++ +  +  +$-1    +     -1    +      -1 * *  *  *  * + * *  * ^ -1         +     : -1        +    $ -1        +     +/!!%   (         DKKr&r) __future__rloggingrvtypingrrrrrr r langchain_core.documentsr langchain_core.embeddingsr langchain_core.vectorstoresr r)rrr getLoggerr,rr$r&r"rsI"LLL-03WW    LK+LKr&