o ZhK@sddlmZddlZddlZddlmZmZmZmZm Z m Z m Z m Z ddl mZddlZddlmZddlmZddlmZddlmZdd lmZerPddlZd&ddZ  d'd(ddZd)ddZd*ddZed d!d"d#Gd$d%d%eZ dS)+) annotationsN) TYPE_CHECKINGAnyCallableDictIterableListOptionalTupleuuid4) deprecated)Document) Embeddings) VectorStore)maximal_marginal_relevance index_namestrtext_keyreturnrcCs||dgdgdS)Ntext)nameZdataType)class properties)rrrr`/var/www/html/lang_env/lib/python3.10/site-packages/langchain_community/vectorstores/weaviate.py_default_schemas rurl Optional[str]api_keykwargsrweaviate.ClientcKspzddl}Wn tytdw|ptjd}|p!tjd}|r+|jj|dnd}|jd||d|S)Nr_Could not import weaviate python package. Please install it with `pip install weaviate-client`Z WEAVIATE_URLZWEAVIATE_API_KEY)r)rZauth_client_secretr)weaviate ImportErrorosenvirongetauthZ AuthApiKeyClient)rrr r#r(rrr_create_weaviate_client)s  r*valfloatcCsdddt|S)N)npexp)r+rrr_default_score_normalizer;sr0valuecCst|tjr |S|SN) isinstancedatetime isoformat)r1rrr_json_serializable?s r6z0.3.18z1.0z&langchain_weaviate.WeaviateVectorStore)ZsinceZremovalZalternative_importc @seZdZdZddedfdIddZedJddZdKddZ dLdMdd Z !dNdOd&d'Z !dNdOd(d)Z !dNdPd+d,Z ! - .dQdRd2d3Z ! - .dQdSd4d5Z !dNdTd7d8Ze dLdddddd9d:ed;dUdCdDZdLdVdGdHZdS)WWeaviatea`Weaviate` vector store. To use, you should have the ``weaviate-client`` python package installed. Example: .. code-block:: python import weaviate from langchain_community.vectorstores import Weaviate client = weaviate.Client(url=os.environ["WEAVIATE_URL"], ...) weaviate = Weaviate(client, index_name, text_key) NTclientrrrr embeddingOptional[Embeddings] attributesOptional[List[str]]relevance_score_fn"Optional[Callable[[float], float]]by_textboolc Cszddl}Wn tytdwt||js!tdt|||_||_||_||_ |j g|_ ||_ ||_ |durD|j |dSdS)z Initialize with Weaviate client.rNz_Could not import weaviate python package. Please install it with `pip install weaviate-client`.z5client should be an instance of weaviate.Client, got )r#r$r3r) ValueErrortype_client _index_name _embedding _text_key _query_attrsr=_by_textextend) selfr8rrr9r;r=r?r#rrr__init__Zs*     zWeaviate.__init__rcCs|jSr2)rErJrrr embeddings|szWeaviate.embeddingsCallable[[float], float]cCs|jr|jStSr2)r=r0rLrrr_select_relevance_score_fns z#Weaviate._select_relevance_score_fntexts Iterable[str] metadatasOptional[List[dict]]r List[str]c Ksddlm}g}d}|jrt|tst|}|j|}|jj_}t|D]Q\}} |j | i} |durC|| D] \} } t | | | <q8|t } d|vrS|d|} n d|vr]|d|} |j | |j| |ri||nd|dd|| q%Wd|S1swY|S)z4Upload texts with metadata (properties) to Weaviate.rget_valid_uuidNuuidsidstenant) data_object class_nameuuidvectorrY) weaviate.utilrVrEr3listembed_documentsrCbatch enumeraterFitemsr6r add_data_objectrDr'append)rJrPrRr rVrXrMrairdata_propertieskeyr+_idrrr add_textss>         zWeaviate.add_textsquerykintList[Document]cKsL|jr |j||fi|S|jdurtd|j|}|j||fi|S)Return docs most similar to query. Args: query: Text to look up documents similar to. k: Number of Documents to return. Defaults to 4. Returns: List of Documents most similar to the query. NzC_embedding cannot be None for similarity_search when _by_text=False)rHsimilarity_search_by_textrErA embed_querysimilarity_search_by_vector)rJrlrmr r9rrrsimilarity_searchs   zWeaviate.similarity_searchc Ksd|gi}|dr|d|d<|jj|j|j}|dr(||d}|dr5||d}|drB||d}|| | }d|vrYt d|dg}|d d |jD]}| |j } |t| |d qd|S) rpconceptssearch_distance certainty where_filterrY additionalerrorsError during query: dataGetZ page_contentmetadata)r'rCrlrDrG with_where with_tenantwith_additionalwith_near_text with_limitdorApoprFrer) rJrlrmr content query_objresultdocsresrrrrrqs$     z"Weaviate.similarity_search_by_text List[float]c Ksd|i}|jj|j|j}|dr||d}|dr(||d}|dr5||d}|| | }d|vrLt d|dg}|dd|jD]}| |j } |t| |d qW|S) z:Look up similar documents by embedding vector in Weaviate.r]rxrYryrzr{r|r}r~)rCrlr'rDrGrrrwith_near_vectorrrrArrFrer) rJr9rmr r]rrrrrrrrrss     z$Weaviate.similarity_search_by_vector?fetch_k lambda_multr,cKs:|jdur |j|}ntd|j|f|||d|S)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. 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. Returns: List of Documents selected by maximal marginal relevance. NzCmax_marginal_relevance_search requires a suitable Embeddings object)rmrr)rErrrA'max_marginal_relevance_search_by_vector)rJrlrmrrr r9rrrmax_marginal_relevance_searchs z&Weaviate.max_marginal_relevance_searchcKsd|i}|jj|j|j}|dr||d}|dr(||d}|d| | }|dd|j} dd| D} t t || ||d} g} | D]} | | |j}| | d | | }| t||d qT| S) aReturn docs selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to query AND diversity among selected documents. Args: embedding: Embedding 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. 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. Returns: List of Documents selected by maximal marginal relevance. r]rxrYr|r}cSsg|]}|ddqS) _additionalr]r).0rrrr HszDWeaviate.max_marginal_relevance_search_by_vector..)rmrrr~)rCrlr'rDrGrrrrrrrr.arrayrrFrer)rJr9rmrrr r]rresultspayloadrMZ mmr_selectedridxrmetarrrr!s.  z0Weaviate.max_marginal_relevance_search_by_vectorList[Tuple[Document, float]]c Ks<|jdur tdd|gi}|dr|d|d<|jj|j|j}|dr1||d}|dr>||d}|j |}|j sYd|i}| | | d}n || | d}d |vrstd |d g} |d d |jD]} | |j} t| d d|} | t| | d| fq~| S)z Return list of documents most similar to the query text and cosine distance in float for each. Lower score represents more similarity. Nz:_embedding cannot be None for similarity_search_with_scorerurvrwrxrYr]rzr{r|r}rr~)rErAr'rCrlrDrGrrrrrHrrrrrrrFr.dotrer) rJrlrmr rrZembedded_queryr]rZdocs_and_scoresrrZscorerrrsimilarity_search_with_scoreUsB       z%Weaviate.similarity_search_with_scorerF)r8 weaviate_urlweaviate_api_key batch_sizerrr?r=rOptional[weaviate.Client]rrrr Optional[int]c szddlmWnty} ztd| d} ~ ww|p!t||d}|r+|jj|d|p3dtj}t|| }|j |sE|j ||rL| |nd}|rXt |dnd}d| vrd| d}n fd d tt|D}|jJ}t|D]9\}}| |i}|dur||D] }|||||<q||}|||d }|dur|||d <|jdi|qy|Wdn1swY|||| f||| | d | S)avConstruct Weaviate wrapper from raw documents. This is a user-friendly interface that: 1. Embeds documents. 2. Creates a new index for the embeddings in the Weaviate instance. 3. Adds the documents to the newly created Weaviate index. This is intended to be a quick way to get started. Args: texts: Texts to add to vector store. embedding: Text embedding model to use. metadatas: Metadata associated with each text. client: weaviate.Client to use. weaviate_url: The Weaviate URL. If using Weaviate Cloud Services get it from the ``Details`` tab. Can be passed in as a named param or by setting the environment variable ``WEAVIATE_URL``. Should not be specified if client is provided. weaviate_api_key: The Weaviate API key. If enabled and using Weaviate Cloud Services, get it from ``Details`` tab. Can be passed in as a named param or by setting the environment variable ``WEAVIATE_API_KEY``. Should not be specified if client is provided. batch_size: Size of batch operations. index_name: Index name. text_key: Key to use for uploading/retrieving text to/from vectorstore. by_text: Whether to search by text or by embedding. relevance_score_fn: Function for converting whatever distance function the vector store uses to a relevance score, which is a normalized similarity score (0 means dissimilar, 1 means similar). kwargs: Additional named parameters to pass to ``Weaviate.__init__()``. Example: .. code-block:: python from langchain_community.embeddings import OpenAIEmbeddings from langchain_community.vectorstores import Weaviate embeddings = OpenAIEmbeddings() weaviate = Weaviate.from_texts( texts, embeddings, weaviate_url="http://localhost:8080" ) rrUr"N)rr)rZ LangChain_rWcsg|]}tqSrr )r_rUrrrsz'Weaviate.from_texts..)r\rZr[r])r9r;r=r?r)r^rVr$r*ra configurer hexrschemaexistsZ create_classr`r_keysrrangelenrbrdflush)clsrPr9rRr8rrrrrr?r=r errMr;rWrarfrrgrhriparamsrrUr from_textssl@      zWeaviate.from_textsrXNonecKs.|durtd|D] }|jjj|dq dS)zUDelete by vector IDs. Args: ids: List of ids to delete. NzNo ids provided to delete.)r\)rArCrZdelete)rJrXr idrrrr s zWeaviate.delete)r8rrrrrr9r:r;r<r=r>r?r@)rr:)rrNr2)rPrQrRrSr rrrT)rk)rlrrmrnr rrro)r9rrmrnr rrro)rkrr) rlrrmrnrrnrr,r rrro) r9rrmrnrrnrr,r rrro)rlrrmrnr rrr)rPrTr9rrRrSr8rrrrrrrrrrrr?r@r=r>r rrr7)rXr<r rrr)__name__ __module__ __qualname____doc__r0rKpropertyrMrOrjrtrqrsrrr classmethodrrrrrrr7EsT "   ,    ' 5 0r7)rrrrrr)NN)rrrrr rrr!)r+r,rr,)r1rrr)! __future__rr4r%typingrrrrrrr r r\r numpyr.Zlangchain_core._apir Zlangchain_core.documentsrZlangchain_core.embeddingsrZlangchain_core.vectorstoresrZ&langchain_community.vectorstores.utilsrr#rr*r0r6r7rrrrs2 (