dhWLSSKJr SSKrSSKrSSKrSSKJrJrJrJ r J r J r J r SSK rSSKJr SSKJr SSKJr SSKJr SSKJr S rS rS rS rS rSrSr"SS\5r g)) annotationsN)AnyDictListOptionalSequenceTupleType)Document) Embeddings)UpsertResponse) VectorStore)overrideHNSWCS langchainilc_textuniqueidcb^\rSrSr\\SSS\RS4SU4Sjjj5rSSjr \SSSjj5r \SSj5r \SSSjj5r \SSj5r SS jrSSS jjr\SS S jj5r\S!S"S jj5r\S!S#S jj5r\\SS$Sjj55r\\S%Sj55r\S&Sj5r\SSj5rS'SjrS(SjrSrU=r$)) ApertureDBNc >SSKJn Jn [T U]"S0UD6 [ R"[5Ul URRU5 X l Xl X0l X@lXPlXplUc&URR#S5 [%S5eSSKJn Jn U "5UlU "UR(5UlUR*R-5 UR15 g![a [S5ef=f![a URR'S5 ef=f![.a URR'S5 ef=f)a&Create a vectorstore backed by ApertureDB A single ApertureDB instance can support many vectorstores, distinguished by 'descriptor_set' name. The descriptor set is created if it does not exist. Different descriptor sets can use different engines and metrics, be supplied by different embedding models, and have different dimensions. See ApertureDB documentation on `AddDescriptorSet` https://docs.aperturedata.io/query_language/Reference/descriptor_commands/desc_set_commands/AddDescriptorSet for more information on the engine and metric options. Args: embeddings (Embeddings): Embeddings object descriptor_set (str, optional): Descriptor set name. Defaults to "langchain". dimensions (Optional[int], optional): Number of dimensions of the embeddings. Defaults to None. engine (str, optional): Engine to use. Defaults to "HNSW" for new descriptorsets. metric (str, optional): Metric to use. Defaults to "CS" for new descriptorsets. log_level (int, optional): Logging level. Defaults to logging.WARN. rUtilscreate_connectorzMApertureDB is not installed. Please install it using 'pip install aperturedb'NzNo embedding function provided.zFailed to connect to ApertureDB)aperturedb.Utilsrr ImportErrorsuper__init__logging getLogger__name__loggersetLeveldescriptor_setembedding_function dimensionsenginemetric propertiesfatal ValueError exception connectionutilsstatus Exception_find_or_add_descriptor_set) self embeddingsr'r)r*r+ log_levelr,kwargsrr __class__s c/var/www/html/shao/venv/lib/python3.13/site-packages/langchain_community/vectorstores/aperturedb.pyr!ApertureDB.__init__sDJ  @ "6"''1  Y',",$  $   KK  ? @>? ?  @+,4??+   JJ    ((*K +  ,  KK ! !+      KK ! !"C D  s#C:#DD=:D'D:='E$c|URnSUSSSSS0S.0/nURRU5up4URR5(dU5eSUSS;a[ USSS5OSnUS::dS5eUS:XGaUSSSSnUR R S US 35 US n[ U5S:XdS 5eURc USUlOAURUS:wa.UR RS URSUS35 USn[ U5S:XdS5eURc USUl OAURUS:wa.UR RSURSUS35 USn URcXl O;URU :wa+UR RSURSU 35 UR5V V s0sH4upU R[5(dM!U [ [5SU _M6 sn n UlgUR R S US35 URc [UlURc [ Ul URc)[ UR"R%S55Ul URb6URR5V V s0sHup[U -U _M sn n OSn UR&R)UURURURU S9 UR&R+SS5 UR&R+SS5 UR&R+S[,5 gs sn n fs sn n f)NFindDescriptorSetTall_properties) with_nameenginesmetricsr)resultsentitiesrz+Multiple descriptor sets with the same namezDescriptor set z already exists_engineszOnly one engine is supportedzEngine mismatch: z != _metricszOnly one metric is supportedzMetric mismatch: _dimensionszDimensions mismatch: z does not exist. Creating ittest)namedimr*r+r, _Descriptor _create_txn_DescriptorSet_name)r'r0query last_query_oklenr%infor*errorr+r)items startswithPROPERTY_PREFIXr,ENGINEMETRICr( embed_queryr1add_descriptorsetcreate_entity_indexUNIQUEID_PROPERTY) r5r' find_ds_queryrb n_entitieser@rAr)kvr,s r:r4&ApertureDB._find_or_add_descriptor_setls,,E$!/##"& 0$7 &   $$]3,,..11.QqT"566 !()*5 6  QM MM ?!()*5a8A KK  ~.>oN O mGw<1$ D&D D${{"%aj  * !!$5dkk]$wqzl"ST mGw<1$ D&D D${{"%aj  * !!$5dkk]$wqzl"ST=)J&",J. !!+DOO+T)rhrirBrrC)r'r\r1rjget_descriptor_to_document)r5rkrOrBrmddocss r: get_by_idsApertureDB.get_by_idss!..$5c{#C 0$7#  ZZ''. QZ 0155j"E E  ( ( +E    sA3cURcS5eURRU5nUR"XR/UQ70UD6$)zSearch for documents similar to the query using the vectorstore Args: query: Query string to search for. k: Number of results to return. Returns: List of Document objects ordered by decreasing similarity to the query. zEmbedding function is not set)r(rYsimilarity_search_by_vector)r5rOrbargsr8 embeddings r:similarity_searchApertureDB.similarity_searchsM&&2S4SS2++77> // NtNvNNrpcdURRU5nUR"U/UQ70UD6$N)r(rY'_similarity_search_with_score_by_vector)r5rOr{r8r|s r:similarity_search_with_score'ApertureDB.similarity_search_with_scores6++77> ;;IWWPVWWrpc0nUR5H5up4UR[5(dM!XBU[[5S'M7 U[nU[ n[ XRUS9nU$)N page_contentmetadataid)rTrUrVrQ TEXT_PROPERTYr\r )r5rurrbrcrrdocs r:rt"ApertureDB._descriptor_to_document seGGIDA||O,,673/123&'DI rpcPSSKJn U"UR5n[R"5nUR UR XSS9 UR RS[R"5U- S35 UVs/sHopRU5US4PM sn$s snf)Nr DescriptorsT)rhvector k_neighbors distances"ApertureDB similarity search took  seconds _distance aperturedb.Descriptorsrr0time find_similarr'r%rRrt)r5r|rbvectorsr descriptors start_timerus r:r2ApertureDB._similarity_search_with_score_by_vectors 7!$//2 YY[   ##IPT !  0z1I0J( S LWW;a--a0!K.A;WWWsB#c DSSKJn U"UR5n[R"5nUR UR XS9 UR RS[R"5U- S35 UVs/sHopRU5PM sn$s snf)aReturns the k most similar documents to the given embedding vector Args: embedding: The embedding vector to search for k: The number of similar documents to return Returns: List of Document objects ordered by decreasing similarity to the query. rr)rhrrrrr)r5r|rbr8rrrrus r:rz&ApertureDB.similarity_search_by_vector%s 7!$//2 YY[   ##I !  0z1I0J( S :EEA,,Q/EEEsBc URRSU35 URRU5nUR"XbX440UD6$)aReturns similar documents to the query that also have diversity This algorithm balances relevance and diversity in the search results. Args: query: Query string to search for. k: Number of results to return. fetch_k: Number of results to fetch. lambda_mult: Lambda multiplier for MMR. Returns: List of Document objects ordered by decreasing similarity/diversty. z)Max Marginal Relevance search for query: )r%rRr(rY'max_marginal_relevance_search_by_vector)r5rOrbfetch_k lambda_multr8r|s r:max_marginal_relevance_search(ApertureDB.max_marginal_relevance_search>sT, DUGLM++77> ;; ' 28  rpc JSSKJn U"UR5n[R"5nUR UR UUUUS9 UR RS[R"5U- S35 UV s/sHoRU 5PM sn $s sn f)aReturns similar documents to the vector that also have diversity This algorithm balances relevance and diversity in the search results. Args: embedding: Embedding vector to search for. k: Number of results to return. fetch_k: Number of results to fetch. lambda_mult: Lambda multiplier for MMR. Returns: List of Document objects ordered by decreasing similarity/diversty. rr)rhrrrrz&ApertureDB similarity search mmr took r) rrr0rfind_similar_mmrr'r%rRrt) r5r|rbrrr8rrrrus r:r2ApertureDB.max_marginal_relevance_search_by_vectorZs, 7!$//2 YY[ $$### %  4TYY[:5M4Nh W :EEA,,Q/EEEsB c <U"SSU0UD6nURX5 U$)aCreates a new vectorstore from a list of texts Args: texts: List of text strings embedding: Embeddings object as for constructing the vectorstore metadatas: Optional list of metadatas associated with the texts. kwargs: Additional arguments to pass to the constructor r6r) add_texts)clstextsr| metadatasr8stores r: from_textsApertureDB.from_textss'"3y3F3 ) rpc <U"SSU0UD6nURU5 U$)a(Creates a new vectorstore from a list of documents Args: documents: List of Document objects embedding: Embeddings object as for constructing the vectorstore metadatas: Optional list of metadatas associated with the texts. kwargs: Additional arguments to pass to the constructor r6r) add_documents)r documentsr|r8rs r:from_documentsApertureDB.from_documentss) 3y3F3 I& rpcTSSKJnJn U"5nU"U5nURU5 g)z}Deletes a vectorstore and all its data from the database Args: descriptor_set: The name of the descriptor set to delete rrN)rrrremove_descriptorset)class_r'rrdbr1s r:delete_vectorstoreApertureDB.delete_vectorstores& =  b  "">2rpcSSKJn U"5nSSS0SSSS.0/nURU5upEUR5(dU5eUSSS$)ziReturns a list of all vectorstores in the database Returns: List of descriptor sets with properties r)rr=r>T)rBr@rAr)rC)rrrOrP)rrrrOresponserms r:list_vectorstoresApertureDB.list_vectorstoressv 6  $ 0$7##"& &   hhuo !!+8+!{./ ;;rpc BSU;aURS5nU(a#[U5[U5:wa [S5e/n[X15H5upV[ UR UR US9nURU5 M7 OUnUR"U40UD6S$)aAdd or update documents in the vectorstore. Args: documents: Documents to add to the vectorstore. kwargs: Additional keyword arguments. if kwargs contains ids and documents contain ids, the ids in the kwargs will receive precedence. Returns: List of IDs of the added texts. Raises: ValueError: If the number of ids does not match the number of documents. rkzhThe number of ids must match the number of documents. Got {len(ids)} ids and {len(documents)} documents.r succeeded) poprQr.zipr rrappendupsert)r5rr8rk documents_id_document doc_with_ids r:rApertureDB.add_documentss F?**U#Cs3x3y>1 I J!$S!4 &!)!6!6%.. !!+. "5#J{{:00==rpc SSKJn UVs/sH1n[US5(dMURcM%URPM3 nnU(aUR U5 UVs/sHofR PM nnUVs/sHn[ USS5b URO0PM! nnURRU5n UVs/sHKn[US5(aURb URO[[R"55PMM n n/n [XyX5HuppUR5VVs0sHunn[U-U_M nnnU U[ 'UU["'SUR$US.0nU/n[&R("U [&R*S9R-5/nU R/UU45 M U"UR05nUR3U [4S 9 [7U /S 9$s snfs snfs snfs snfs snnf) zInsert or update items Updating documents is dependent on the documents' `id` attribute. Args: items: List of Document objects to upsert Returns: UpsertResponse object with succeeded and failed r)ParallelLoaderrNr AddDescriptor)rhr,)dtype) batchsize)rfailed)aperturedb.ParallelLoaderrhasattrrrnrgetattrrr(embed_documentsstruuiduuid4rrTrVrr\r'nparrayfloat32tobytesrr0ingest BATCHSIZEr )r5rTr8ritem ids_to_deleterrrr6rkdatarr|r unique_idrbrcr,commandrOblobsloaders r:rApertureDB.upserts =!&$ %t)F^^=MN=MTQ/A-q0=MJN(,J} %,5J( )..","G IEXXirzz:BBDEE KK '5  0 di 0B77G$ 4  Os(G2 G2G2 G7:%G<AH:H) r0r'r)r(r*r%r+r,r1)r6r r'rr)z Optional[int]r* Optional[str]r+rr7intr,zOptional[Dict]r8rreturnNone)rrr)rkzOptional[List[str]]r8rrzOptional[bool])rkz Sequence[str]rList[Document])) rOrrbrr{rr8rrr)rOrr{rr8rrList[Tuple[Document, float]])rudictrr )rF)r| List[float]rbrrboolrr)r|rrbrr8rrr)rg?) rOrrbrrrrfloatr8rrr) r|rrbrrrrrr8rrr) rType[ApertureDB]r List[str]r|r rzOptional[List[dict]]r8rrr) rrrrr|r r8rrr)r'rrr)rrr8rrr)rTzSequence[Document]r8rrr )r$ __module__ __qualname____firstlineno__rDESCRIPTOR_SETr"WARNr!r4rnrwr}rrtrrzrr classmethodrrrrrr__static_attributes__ __classcell__)r9s@r:rrsJ -$( $ $ %)K+K+K+" K+  K+  K+K+#K+K+ K+K+ZTMl,4#$OO O-0ON4848rpr)! __future__rr"rrtypingrrrrrr r numpyrlangchain_core.documentsr langchain_core.embeddingsr langchain_core.indexing.baser langchain_core.vectorstoresrtyping_extensionsrrWrXrrrVrr\rrrpr:rsd" CCC.073&     M8M8rp