3fiztdZddlmZddlZddlZddlZddlmZddlm Z m Z m Z m Z m Z mZmZmZmZmZddlmZddlmZddlmZdd lmZdd lmZe rdd lmZej>Z d Z!d Z"GddeZ#y)z?VectorStore wrapper around a Postgres-TimescaleVector database.) annotationsN) timedelta) TYPE_CHECKINGAnyCallableDictIterableListOptionalTupleTypeUnion)Document) Embeddings)get_from_dict_or_env) VectorStore)DistanceStrategy) Predicatesilangchain_storeceZdZdZeeeddddf d&dZ d'dZe d(dZ d'dZ e ddeeddf d)dZ e ddeeddf d)d Z d* d+d Z d* d+d Z d* d,d Z d* d,d Zd-dZ d. d/dZ d. d/dZ d. d0dZ d. d0dZd1dZ d. d2dZ d. d2dZ d. d3dZ d. d3dZe deeddf d4dZe deeddf d4dZe deeddf d5dZe deeddf d5dZ e eedf d6dZ!e d7dZ"e d8dZ#d9dZ$d:d;d Z% dTimescaleVectoraTimescale Postgres vector store To use, you should have the ``timescale_vector`` python package installed. Args: service_url: Service url on timescale cloud. embedding: Any embedding function implementing `langchain.embeddings.base.Embeddings` interface. collection_name: The name of the collection to use. (default: langchain_store) This will become the table name used for the collection. distance_strategy: The distance strategy to use. (default: COSINE) pre_delete_collection: If True, will delete the collection if it exists. (default: False). Useful for testing. Example: .. code-block:: python from langchain_community.vectorstores import TimescaleVector from langchain_community.embeddings.openai import OpenAIEmbeddings SERVICE_URL = "postgres://tsdbadmin:@.tsdb.cloud.timescale.com:/tsdb?sslmode=require" COLLECTION_NAME = "state_of_the_union_test" embeddings = OpenAIEmbeddings() vectorestore = TimescaleVector.from_documents( embedding=embeddings, documents=docs, collection_name=COLLECTION_NAME, service_url=SERVICE_URL, ) FNc   ddlm} ||_||_||_||_||_||_|xstjt|_ ||_ | |_| j|j|j |j |jj j#fd|ji| |_| j&|j|j |j |jj j#fd|ji| |_|j+y#t$r tdwxYw)NrclienthCould not import timescale_vector python package. Please install it with `pip install timescale-vector`.time_partition_interval)timescale_vectorr ImportError service_url embeddingcollection_namenum_dimensions_distance_strategypre_delete_collectionlogging getLogger__name__loggeroverride_relevance_score_fn_time_partition_intervalSyncvaluelower sync_clientAsync async_client __post_init__) selfrr r!r"distance_strategyr$r(relevance_score_fnrkwargsrs n/var/www/auto_recruiter/arenv/lib/python3.12/site-packages/langchain_community/vectorstores/timescalevector.py__init__zTimescaleVector.__init__HsV  /'".,"3%:"; 1 1( ; +=((?%&6;;         # # ) ) / / 1  %)$A$A    )FLL         # # ) ) / / 1  %)$A$A     ? I  s EEc|jj|jr|jjyy)z' Initialize the store. N)r. create_tablesr$ delete_allr2s r6r1zTimescaleVector.__post_init__xs6 &&(  % %    ' ' ) &c|jSN)r r;s r6 embeddingszTimescaleVector.embeddingss ~~r<c8|jjyr>)r. drop_tabler;s r6 drop_tableszTimescaleVector.drop_tabless ##%r<c "t|d} |*|D cgc]} ttj!}} |s|D cgc]} i}} ||j | }|d|| |||| d| } | j d||||d| | Scc} wcc} wNr)rr"r!r r3r$textsr? metadatasids)lenstruuiduuid4get_service_urladd_embeddingsclsrFr?r rGrHr!r3rr$r5r"_stores r6__fromzTimescaleVector.__fromsZ]+ ;.343tzz|$4C4%*++I+  --f5K #)+/"7     J) PV  -5,s $B B c >Kt|d} |*|D cgc]} ttj!}} |s|D cgc]} i}} ||j | }|d|| |||| d| } | j d||||d| d{| Scc} wcc} w7wrD)rJrKrLrMrNaadd_embeddingsrPs r6__afromzTimescaleVector.__afromsZ]+ ;.343tzz|$4C4%*++I+  --f5K #)+/"7   $e## J) PV    -5, s'B$BB B >B B Bc |*|Dcgc]}ttj!}}|s|Dcgc]}i}}tt ||||}|j j ||Scc}wcc}w)/Add embeddings to the vectorstore. Args: texts: Iterable of strings to add to the vectorstore. embeddings: List of list of embedding vectors. metadatas: List of metadatas associated with the texts. kwargs: vectorstore specific parameters )rKrLrMlistzipr.upsertr2rFr?rGrHr5rRrecordss r6rOzTimescaleVector.add_embeddingssu ;.343tzz|$4C4%*++I+s3 5*=> ( 5,s $A2 A7cK|*|Dcgc]}ttj!}}|s|Dcgc]}i}}tt ||||}|j j |d{|Scc}wcc}w7w)rYN)rKrLrMrZr[r0r\r]s r6rVzTimescaleVector.aadd_embeddingss ;.343tzz|$4C4%*++I+s3 5*=>&&w/// 5, 0s%B$A<B B7B5B6Bc x|jjt|}|jd||||d|S)rRun 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. kwargs: vectorstore specific parameters Returns: List of ids from adding the texts into the vectorstore. rErI)r embed_documentsrZrOr2rFrGrHr5r?s r6 add_textszTimescaleVector.add_textssG"^^33DK@ "t"" J) PV  r<cK|jjt|}|jd||||d|d{S7w)rarENrI)r rbrZrVrcs r6 aadd_textszTimescaleVector.aadd_texts'sU"^^33DK@ )T)) J) PV    s?AAAch||dk(s|jry|jj|S)N)isspacer embed_query)r2querys r6 _embed_queryzTimescaleVector._embed_query=s- =ERK5==?>>--e4 4r<c R|j|}|jd||||d|S)eRun similarity search with TimescaleVector with distance. Args: query (str): Query text to search for. k (int): Number of results to return. Defaults to 4. filter (Optional[Dict[str, str]]): Filter by metadata. Defaults to None. Returns: List of Documents most similar to the query. r kfilter predicatesrI)rlsimilarity_search_by_vectorr2rkrprqrrr5r s r6similarity_searchz!TimescaleVector.similarity_searchDsD$%%e, /t// !     r<cnK|j|}|jd||||d|d{S7w)rnroNrI)rlasimilarity_search_by_vectorrts r6asimilarity_searchz"TimescaleVector.asimilarity_search_sR$%%e, 6T66 !       ,535c V|j|}|jd||||d|}|S)bReturn docs most similar to query. Args: query: Text to look up documents similar to. k: Number of Documents to return. Defaults to 4. filter (Optional[Dict[str, str]]): Filter by metadata. Defaults to None. Returns: List of Documents most similar to the query and score for each rorI)rl&similarity_search_with_score_by_vector)r2rkrprqrrr5r docss r6similarity_search_with_scorez,TimescaleVector.similarity_search_with_scorezsH$%%e, :t:: !     r<cnK|j|}|jd||||d|d{S7w)r{roNrI)rl'asimilarity_search_with_score_by_vectorrts r6asimilarity_search_with_scorez-TimescaleVector.asimilarity_search_with_scoresR&%%e, ATAA !       ryc dDcic] }||vr|||}}|rt|dk(ry ddlm}|jdi|Scc}w#t$r tdwxYw)N) start_dateend_date time_deltastart_inclusive end_inclusiverrrrI)rJrrr UUIDTimeRange)r2r5keyconstructor_argsrs r6date_to_range_filterz$TimescaleVector.date_to_range_filters  f}       3'7#8A#=  / $v##7&677+   I  sAAAc 4 ddlm}|jj |||||j di|}|Dcgc]8}t ||j||j||jf:} }| S#t$r tdwxYwcc}wNrrr)limitrqrruuid_time_filter) page_contentmetadatarI) rrrr.searchrrSEARCH_RESULT_CONTENTS_IDXSEARCH_RESULT_METADATA_IDXSEARCH_RESULT_DISTANCE_IDX r2r rprqrrr5rresultsresultr}s r6r|z6TimescaleVector.similarity_search_with_score_by_vectors  /"")) !6T66@@ *  "   !'(I(I!J#F$E$EFv889       1 I    sA==B=Bc PK ddlm}|jj |||||j di|d{}|Dcgc]8}t ||j||j||jf:} }| S#t$r tdwxYw7acc}wwr) rrrr0rrrrrrrs r6rz7TimescaleVector.asimilarity_search_with_score_by_vectors  /))00 !6T66@@ 1   "   !'(I(I!J#F$E$EFv889       1 I      s7B&B3B&BB&=B!B&BB&!B&c f|jd||||d|}|Dcgc]\}}| c}}Scc}}w)kReturn docs most similar to embedding vector. Args: embedding: Embedding to look up documents similar to. k: Number of Documents to return. Defaults to 4. filter (Optional[Dict[str, str]]): Filter by metadata. Defaults to None. Returns: List of Documents most similar to the query vector. rorI)r| r2r rprqrrr5docs_and_scoresdocrRs r6rsz+TimescaleVector.similarity_search_by_vectorsG$F$EE 1V  NT #22Q222s -cK|jd||||d|d{}|Dcgc]\}}| c}}S7cc}}ww)rroNrI)rrs r6rwz,TimescaleVector.asimilarity_search_by_vector)sW$!M L L! 1V ! NT!  #22Q22 3s?7 ? 9??c l|jt|} |j|| |f|||||d|S) Return VectorStore initialized from texts and embeddings. Postgres connection string is required "Either pass it as a parameter or set the TIMESCALE_SERVICE_URL environment variable. rGrHr!r3r$)rbrZ_TimescaleVector__from rQrFr rGr!r3rHr$r5r?s r6 from_textszTimescaleVector.from_texts@sT$..tE{; szz      +/"7    r<c K|jt|} |j|| |f|||||d|d{S7w)rrN)rbrZ_TimescaleVector__afromrs r6 afrom_textszTimescaleVector.afrom_texts`sb$..tE{;  S[[      +/"7      s9AAAc |D cgc]} | d } } |D cgc]} | d } } |j| | |f|||||d|Scc} wcc} w)$Construct TimescaleVector wrapper from raw documents and pre- generated embeddings. Return VectorStore initialized from documents and embeddings. Postgres connection string is required "Either pass it as a parameter or set the TIMESCALE_SERVICE_URL environment variable. Example: .. code-block:: python from langchain_community.vectorstores import TimescaleVector from langchain_community.embeddings import OpenAIEmbeddings embeddings = OpenAIEmbeddings() text_embeddings = embeddings.embed_documents(texts) text_embedding_pairs = list(zip(texts, text_embeddings)) tvs = TimescaleVector.from_embeddings(text_embedding_pairs, embeddings) rr)r rQtext_embeddingsr rGr!r3rHr$r5trFr?s r6from_embeddingszTimescaleVector.from_embeddingssv< //!1//$34qad4 4szz      +/"7    04s A Ac K|D cgc]} | d } } |D cgc]} | d } } |j| | |f|||||d|d{Scc} wcc} w7w)rrrrN)rrs r6afrom_embeddingsz TimescaleVector.afrom_embeddingss< //!1//$34qad4 4 S[[      +/"7     04  s%A A A A AAAc B|j|}||||||}|S)z Get instance of an existing TimescaleVector store.This method will return the instance of the store without inserting any new embeddings )rr!r r3r$)rN)rQr r!r3r$r5rrSs r6from_existing_indexz#TimescaleVector.from_existing_indexs4))&1 #+/"7   r<c<t|dd}|s td|S)NrTIMESCALE_SERVICE_URL)datarenv_keyzyPostgres connection string is requiredEither pass it as a parameteror set the TIMESCALE_SERVICE_URL environment variable.)r ValueError)rQr5rs r6rNzTimescaleVector.get_service_urls5/+ I  r<c $d|d|d|d|d| S)z2Return connection string from database parameters.z postgresql://:@/rI)rQhostportdatabaseuserpasswords r6service_url_from_db_paramsz*TimescaleVector.service_url_from_db_paramss)tfAhZqavQxjIIr<cZ|j |jS|jtjk(r |jS|jtj k(r |j S|jtjk(r |jStd|jd)a8 The 'correct' relevance function may differ depending on a few things, including: - the distance / similarity metric used by the VectorStore - the scale of your embeddings (OpenAI's are unit normed. Many others are not!) - embedding dimensionality - etc. z=No supported normalization function for distance_strategy of zF.Consider providing relevance_score_fn to TimescaleVector constructor.) r)r#rCOSINE_cosine_relevance_score_fnEUCLIDEAN_DISTANCE_euclidean_relevance_score_fnMAX_INNER_PRODUCT%_max_inner_product_relevance_score_fnrr;s r6_select_relevance_score_fnz*TimescaleVector._select_relevance_score_fns  + + 733 3  " "&6&=&= =22 2  $ $(8(K(K K55 5  $ $(8(J(J J== =--1-D-D,EFXX r<c T| td|jj|y)3Delete by vector ID or other criteria. Args: ids: List of ids to delete. **kwargs: Other keyword arguments that subclasses might use. Returns: Optional[bool]: True if deletion is successful, False otherwise, None if not implemented. zNo ids provided to delete.T)rr. delete_by_ids)r2rHr5s r6deletezTimescaleVector.delete-s, ;9: : &&s+r<c :|jj|y)rT)r.delete_by_metadata)r2rqr5s r6rz"TimescaleVector.delete_by_metadata?s ++F3r<ceZdZdZdZdZdZy)TimescaleVector.IndexTypez(Enumerator for the supported Index typestsvivfflathnswN)r' __module__ __qualname____doc__TIMESCALE_VECTORPGVECTOR_IVFFLAT PGVECTOR_HNSWrIr<r6 IndexTyperPs6 $ r<rc ` ddlm}t||jr |j n|}||jj j k(r+|jj|jdi|||jjj k(r+|jj|jdi|||jjj k(r,|jj|jdi|yy#t$r tdwxYw)NrrrrI)rrr isinstancerr,rr.create_embedding_index IvfflatIndexr HNSWIndexrTimescaleVectorIndex)r2 index_typer5rs r6 create_indexzTimescaleVector.create_indexYs   /!+:t~~ FJ  J  88>> >    3 34GF4G4G4Q&4Q R 55;; ;    3 34DF4D4D4Nv4N O 88>> >    3 3+++5f5  ? I  s DD-c8|jjyr>)r.drop_embedding_indexr;s r6 drop_indexzTimescaleVector.drop_indexrs --/r<)rrKr rr!rKr"intr3rr$boolr(zOptional[logging.Logger]r4z"Optional[Callable[[float], float]]rzOptional[timedelta]r5rreturnNone)rr)rr)rF List[str]r?List[List[float]]r rrGOptional[List[dict]]rHOptional[List[str]]r!rKr3rrz Optional[str]r$rr5rrr)NN) rF Iterable[str]r?rrGrrHrr5rrr) rFrrGrrHrr5rrr)rkrKrOptional[List[float]])NN) rkrKrprrqOptional[Union[dict, list]]rrOptional[Predicates]r5rrList[Document]) rkrKrprrqrrrrr5rrList[Tuple[Document, float]])r5rrr) r rrprrqrrrrr5rrr) r rrprrqrrrrr5rrr)rQType[TimescaleVector]rFrr rrGrr!rKr3rrHrr$rr5rrr)rzList[Tuple[str, List[float]]]r rrGrr!rKr3rrHrr$rr5rrr)rQrr rr!rKr3rr$rr5rrr)r5zDict[str, Any]rrK) rrKrrrrKrrKrrKrrK)rzCallable[[float], float]r>)rHrr5rrOptional[bool])rqz+Union[Dict[str, str], List[Dict[str, str]]]r5rrr)rzUnion[IndexType, str]r5rrr)/r'rrr"_LANGCHAIN_DEFAULT_COLLECTION_NAMEADA_TOKEN_COUNTDEFAULT_DISTANCE_STRATEGYr7r1propertyr?rB classmethodrrrOrVrdrfrlrurxr~rrr|rrsrwrrrrrrNrrrrrKenumEnumrrDEFAULT_INDEX_TYPErrrIr<r6rr(sF B-.G&++/AE7;... .  . , . $.).?."5.. .`* *& +/#'A.G%)&+&&&& & ( & ! &&,&#& $&& &&P +/#'A.G%)&+&&&& & ( & ! &&,&#& $&& &&X+/#' &(  !    >+/#' &(  !    <+/#'   ( !      2+/#'   ( !      ,5.2+/     ,  )      <.2+/     ,  )      <.2+/  ,  )    &>.2+/     ,  )     & 886.2+/ "(" ", " ) "  " &"N.2+/ "(" ", " ) "  " &"N.2+/ 3(3 3, 3 ) 3  3 34.2+/ 3(3 3, 3 ) 3  3 3. +/A.G#'&+ "   (    ,  !  $     > +/A.G#'&+ "   (    ,  !  $     > +/A.G#'&+* 6* * ( *  * , * !*  $* *  * * X +/A.G#'&+* 6* * ( *  * , * !*  $* *  * * X B.G&+  ",  $    4  J J J J  J  J  J J6$AMP "C#333E/PS 20r<r)$r __future__rrr%rLdatetimertypingrrrrr r r r r rlangchain_core.documentsrlangchain_core.embeddingsrlangchain_core.utilsrlangchain_core.vectorstoresr&langchain_community.vectorstores.utilsrrrrrrrrrIr<r6r sdE"     .053C+-33%6"K 0kK 0r<