NgHddlmZddlZddlZddlZddlmZmZmZm Z m Z m Z m Z m Z mZddlZddlmZddlmZddlmZddlmZddlmZdd lmZerddlZddlZdd lm Z m!Z!m"Z"m#Z#ej$Z%d Z&ddZ'ddZ(edddGddeZ)dS)) annotationsN) TYPE_CHECKINGAnyCallableDictIterableListOptionalTupleType) deprecatedDocument) Embeddings)xor_args) VectorStore)maximal_marginal_relevance)ID OneOrManyWhere WhereDocumentresultsrreturnList[Document]c4dt|DS)Ncg|]\}}|Sr.0doc_s c/var/www/html/ai-engine/env/lib/python3.11/site-packages/langchain_community/vectorstores/chroma.py z$_results_to_docs..%s C C CFCC C C C)_results_to_docs_and_scoresrs r#_results_to_docsr($s C C9'BB C C CCr%List[Tuple[Document, float]]cdt|dd|dd|ddDS)Nc\g|])}t|d|dpi|df*S)r) page_contentmetadatar)r results r#r$z/_results_to_docs_and_scores..)sL     vay6!9? C C CVAYO   r% documentsr metadatas distances)zipr's r#r&r&(sV   K  # K  # K  #     r%z0.2.91.0zlangchain_chroma.Chroma)sinceremovalalternative_importc eZdZUdZdZded<eddddddfdadZedbdZe d dcddd%Z ded'Z dfdgd-Z dfdhd0Z edfdid4Zeddfdjd7Zeddfdkd9Zeddfdld:Zdmd<Zedfdnd=Zedfdod>Zed?d@ddfdpdDZed?d@ddfdqdEZdrdFZ dsdtdOZedPdQdRSdrdTZdudXZdvdZZedddeddddfdwd]Zeddeddddfdxd^Z dydzd_Z!d{d`Z"dS)|Chromaa`ChromaDB` vector store. To use, you should have the ``chromadb`` python package installed. Example: .. code-block:: python from langchain_community.vectorstores import Chroma from langchain_community.embeddings.openai import OpenAIEmbeddings embeddings = OpenAIEmbeddings() vectorstore = Chroma("langchain_store", embeddings) langchainstr"_LANGCHAIN_DEFAULT_COLLECTION_NAMENcollection_nameembedding_functionOptional[Embeddings]persist_directory Optional[str]client_settings"Optional[chromadb.config.Settings]collection_metadataOptional[Dict]clientOptional[chromadb.Client]relevance_score_fn"Optional[Callable[[float], float]]rNonec@ ddl}ddl}n#t$rtdwxYw|||_||_||_n-|rc|p|j|_|jK|jd\} } } t| dkrt| dkrd|_ |} n|r|jd\} } } t| dkr/t| dkr|j d} n|j d } || _n|j } | |_|j | |_| jp||_||_|j|d| |_||_dS) z Initialize with a Chroma client.rNXCould not import chromadb python package. Please install it with `pip install chromadb`..rzduckdb+parquet)chroma_db_implT) is_persistent)namer?r.)chromadbchromadb.config ImportError_client_settings_client_persist_directoryrA __version__splitintrOconfigSettingsClient_embedding_functionget_or_create_collection _collectionoverride_relevance_score_fn) selfr>r?rArCrErGrIrRmajorminorr"rUs r#__init__zChroma.__init__Gs  OOO " " " " "   A    $3D !!DL&7D # # >&J)J 1#4@&.&:&@&@&E&EOE5!5zzQ3u::>>9I6#2  " >"*"6"<"(((s %c|jSN)r^rbs r# embeddingszChroma.embeddingss ''r%) query_textsquery_embeddingsrrjOptional[List[str]]rkOptional[List[List[float]]] n_resultsrZwhereOptional[Dict[str, str]]where_documentkwargsrrc x ddl}n#t$rtdwxYw|jjd|||||d|S)zQuery the chroma collection.rNrM)rjrkrnrorqr)rRrTr`query)rbrjrkrnrorqrrrRs r#__query_collectionzChroma.__query_collections  OOOO   A   &t% #-)       s!urict|d5}tj|dcdddS#1swxYwYdS)z!Get base64 string from image URI.rbzutf-8N)openbase64 b64encodereaddecode)rbrv image_files r# encode_imagezChroma.encode_images #t__ G #JOO$5$566==gFF G G G G G G G G G G G G G G G G G Gs9AAAuris List[str]r2Optional[List[dict]]idsc fd|D d|Ddj0tjdrj|rct|tz }|r ig|zzg}g}t D]2\}} | r||||3|rfd|Dfd|D} rfd|Dnd} fd |D} j| | | nG#t$r:} d t| vr"d }t| j d dz|z| d} ~ wwxYw|rKfd|D}rfd|Dnd}fd|D}j|||njS)asRun more images through the embeddings and add to the vectorstore. Args: uris List[str]: File path to the image. metadatas (Optional[List[dict]], optional): Optional list of metadatas. ids (Optional[List[str]], optional): Optional list of IDs. Returns: List[str]: List of IDs of the added images. c<g|]}|S))rv)r)r rvrbs r#r$z%Chroma.add_images..s*@@@CT&&3&//@@@r%NcNg|]"}ttj#Srr<uuiduuid4r r"s r#r$z%Chroma.add_images..s&3333tz||$$333r% embed_imagerc g|] }| Srrr idxr2s r#r$z%Chroma.add_images..EEEYs^EEEr%c g|] }| Srr)r r b64_textss r#r$z%Chroma.add_images..s(Q(Q(QC3(Q(Q(Qr%c g|] }| Srrr rris r#r$z%Chroma.add_images..>>>Z_>>>r%c g|] }| Srrr rrs r#r$z%Chroma.add_images..$G$G$G#SX$G$G$Gr%r2rir1rExpected metadata value to bezdTry filtering complex metadata using langchain_community.vectorstores.utils.filter_complex_metadata.r c g|] }| Srr)r jrs r#r$z%Chroma.add_images..s+L+L+LQIaL+L+L+Lr%c g|] }| Srrr rris r#r$z%Chroma.add_images..666qZ]666r%c g|] }| Srrr rrs r#r$z%Chroma.add_images..(C(C(CAQ(C(C(Cr%rir1r) r^hasattrrlen enumerateappendr`upsert ValueErrorr<args)rbrr2rrr length_diff empty_ids non_empty_idsrmimages_with_metadatasembeddings_with_metadatasids_with_metadataemsgimages_without_metadatasembeddings_without_metadatasids_without_metadatasrris` `` @@r# add_imageszChroma.add_imagess$A@@@4@@@ ;33d333C  # /G  $m5 5 /1==4=HHJ 4 d))c)nn4K ;%{(:: IM#I.. * *Q*!((----$$S)))) EEEE}EEE (Q(Q(Q(Q=(Q(Q(Q%BLV>>>> >>>>RV*%H$G$G$G$G$G$G! $++"+#<"7- , "   6#a&&@@^)V);c)ABBB  +L+L+L+L)+L+L+L(:DN6666I6666$-)D(C(C(C(C(C(C% '';6-(   # #%# $     s D,, E065E++E0texts Iterable[str]c  dDdt|j|jrcttz }|r ig|zzg}g}t D]2\}} | r||||3|rfd|Dfd|D} rfd|Dnd} fd|D} |j| | | nG#t$r:} dt| vr"d }t| j d d z|z| d} ~ wwxYw|rKfd |D}rfd |Dnd}fd|D}|j|||n|jS)aRun more texts through the embeddings and add to the vectorstore. Args: texts (Iterable[str]): Texts to add to the vectorstore. metadatas (Optional[List[dict]], optional): Optional list of metadatas. ids (Optional[List[str]], optional): Optional list of IDs. Returns: List[str]: List of IDs of the added texts. NcNg|]"}ttj#Srrrs r#r$z$Chroma.add_texts..&4443tz||$$444r%c g|] }| Srrrs r#r$z$Chroma.add_texts..$rr%c g|] }| Srr)r rrs r#r$z$Chroma.add_texts..%s'L'L'Lsc 'L'L'Lr%c g|] }| Srrrs r#r$z$Chroma.add_texts..'rr%c g|] }| Srrrs r#r$z$Chroma.add_texts..)rr%rrzvTry filtering complex metadata from the document using langchain_community.vectorstores.utils.filter_complex_metadata.rrc g|] }| Srr)r rrs r#r$z$Chroma.add_texts..;s*G*G*G58*G*G*Gr%c g|] }| Srrrs r#r$z$Chroma.add_texts..=rr%c g|] }| Srrrs r#r$z$Chroma.add_texts..?rr%r) listr^embed_documentsrrrr`rrr<r)rbrr2rrrrrrrrtexts_with_metadatasrrrrtexts_without_metadatasrrris ``` @r# add_textszChroma.add_textss$ ;44e444C U   # /1AA%HHJ 4 e**s9~~5K ;%{(:: IM#I.. * *Q*!((----$$S)))) EEEE}EEE 'L'L'L'Lm'L'L'L$BLV>>>> >>>>RV*%H$G$G$G$G$G$G! $++"+#<"6- , "   6#a&&@@^)V);c)ABBB  *G*G*G*GY*G*G*G':DN6666I6666$-)D(C(C(C(C(C(C% '';5-(   # #% $     s7D E 5EErtkfilterc <|j||fd|i|}d|DS)acRun similarity search with Chroma. 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[Document]: List of documents most similar to the query text. rcg|]\}}|Srrrs r#r$z,Chroma.similarity_search..as222Q222r%)similarity_search_with_score)rbrtrrrrdocs_and_scoress r#similarity_searchzChroma.similarity_searchMsH"<$; 1  # '-  32/2222r% embedding List[float]c D|jd||||d|}t|S)a}Return docs most similar to embedding vector. Args: embedding (List[float]): Embedding to look up documents similar to. k (int): 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. rkrnrorqr)_Chroma__query_collectionr(rbrrrrqrrrs r#similarity_search_by_vectorz"Chroma.similarity_search_by_vectorcsH *$) &)       (((r%r)c D|jd||||d|}t|S)a Return docs most similar to embedding vector and similarity score. Args: embedding (List[float]): Embedding to look up documents similar to. k (int): Number of Documents to return. Defaults to 4. filter (Optional[Dict[str, str]]): Filter by metadata. Defaults to None. Returns: List[Tuple[Document, float]]: List of documents most similar to the query text and cosine distance in float for each. Lower score represents more similarity. rr)rr&rs r#1similarity_search_by_vector_with_relevance_scoresz8Chroma.similarity_search_by_vector_with_relevance_scores|sH**$) &)      +7333r%c |j|jd|g|||d|}n-|j|}|jd|g|||d|}t|S)aRun similarity search with Chroma 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[Tuple[Document, float]]: List of documents most similar to the query text and cosine distance in float for each. Lower score represents more similarity. N)rjrnrorqrr)r^r embed_queryr&)rbrtrrrqrrrquery_embeddings r#rz#Chroma.similarity_search_with_scores(  # +-d-"G-   GG#6BB5IIO-d-"1!2-   G+7333r%Callable[[float], float]c|jr|jSd}d}|jj}|r ||vr||}|dkr|jS|dkr|jS|dkr|jSt d|d)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. l2z hnsw:spacecosineipzANo supported normalization function for distance metric of type: z=.Consider providing relevance_score_fn to Chroma constructor.)rar`r._cosine_relevance_score_fn_euclidean_relevance_score_fn%_max_inner_product_relevance_score_fnr)rbdistance distance_keyr.s r#_select_relevance_score_fnz!Chroma._select_relevance_score_fns  + 43 3# #,  . 00 -H x  2 2   5 5   = =O19OOO r%c |jt|jdstd|j|g}|jd|||d|}|S)aSearch for similar images based on the given image URI. Args: uri (str): URI of the image to search for. k (int, optional): Number of results to return. Defaults to DEFAULT_K. filter (Optional[Dict[str, str]], optional): Filter by metadata. **kwargs (Any): Additional arguments to pass to function. Returns: List of Images most similar to the provided image. Each element in list is a Langchain Document Object. The page content is b64 encoded image, metadata is default or as defined by user. Raises: ValueError: If the embedding function does not support image embeddings. Nr4The embedding function must support image embedding.rrrrr)r^rrrrrbrvrrrrimage_embeddingrs r#similarity_search_by_imagez!Chroma.similarity_search_by_images0  # +7  $m4 4 +STT T2>>SE>JJ3$2 %     r%c |jt|jdstd|j|g}|jd|||d|}|S)a!Search for similar images based on the given image URI. Args: uri (str): URI of the image to search for. k (int, optional): Number of results to return. Defaults to DEFAULT_K. filter (Optional[Dict[str, str]], optional): Filter by metadata. **kwargs (Any): Additional arguments to pass to function. Returns: List[Tuple[Document, float]]: List of tuples containing documents similar to the query image and their similarity scores. 0th element in each tuple is a Langchain Document Object. The page content is b64 encoded img, metadata is default or defined by user. Raises: ValueError: If the embedding function does not support image embeddings. Nrrrrr)r^rrrrrs r#/similarity_search_by_image_with_relevance_scorez6Chroma.similarity_search_by_image_with_relevance_score s2  # +7  $m4 4 +STT T2>>SE>JJI$H %     r%g?fetch_k lambda_multfloatc  |jd||||gdd|}ttj|tj|dd|| t |} fdt | D} | S) aHReturn 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. filter (Optional[Dict[str, str]]): Filter by metadata. Defaults to None. Returns: List of Documents selected by maximal marginal relevance. )r2r1r3ri)rkrnrorqinclude)dtyperir)rrc"g|] \}}|v | Srr)r ir mmr_selecteds r#r$zBChroma.max_marginal_relevance_search_by_vector..fs'UUU$!Q1 CTCTACTCTCTr%r)rrnparrayfloat32r(r) rbrrrrrrqrrr candidatesselected_resultsrs @r#'max_marginal_relevance_search_by_vectorz.Chroma.max_marginal_relevance_search_by_vector9s8*$) &)III      2 HYbj 1 1 1 L !! $#    &g.. UUUU)J*?*?UUUr%c |jtd|j|}|||||||} | S)a?Return docs selected using the maximal marginal relevance. Maximal marginal relevance optimizes for similarity to query AND diversity among selected documents. Args: query: Text to look up documents similar to. k: Number of Documents to return. Defaults to 4. fetch_k: Number of Documents to fetch to pass to MMR algorithm. lambda_mult: Number between 0 and 1 that determines the degree of diversity among the results with 0 corresponding to maximum diversity and 1 to minimum diversity. Defaults to 0.5. filter (Optional[Dict[str, str]]): Filter by metadata. Defaults to None. Returns: List of Documents selected by maximal marginal relevance. NzBFor MMR search, you must specify an embedding function oncreation.)rrrq)r^rrr) rbrtrrrrrqrrrdocss r#max_marginal_relevance_searchz$Chroma.max_marginal_relevance_searchism6  # +W ,88?? ;;  #) <   r%cN|j|jjdS)zDelete the collection.N)rVdelete_collectionr`rQrhs r#rzChroma.delete_collections$ &&t'7'<=====r%Optional[OneOrMany[ID]]Optional[Where]limit Optional[int]offsetOptional[WhereDocument]rDict[str, Any]cD|||||d}|||d<|jjdi|S)aRGets the collection. Args: ids: The ids of the embeddings to get. Optional. where: A Where type dict used to filter results by. E.g. `{"color" : "red", "price": 4.20}`. Optional. limit: The number of documents to return. Optional. offset: The offset to start returning results from. Useful for paging results with limit. Optional. where_document: A WhereDocument type dict used to filter by the documents. E.g. `{$contains: "hello"}`. Optional. include: A list of what to include in the results. Can contain `"embeddings"`, `"metadatas"`, `"documents"`. Ids are always included. Defaults to `["metadatas", "documents"]`. Optional. )rrorr rqNrr)r`get)rbrrorr rqrrrs r#rz Chroma.getsM4,      'F9 #t#--f---r%z0.1.17zlSince Chroma 0.4.x the manual persistence method is no longer supported as docs are automatically persisted.r5)r6messager7c|jtdddl}|jd\}}}t |dkr.t |dkr|jdSdSdS)a+Persist the collection. This can be used to explicitly persist the data to disk. It will also be called automatically when the object is destroyed. Since Chroma 0.4.x the manual persistence method is no longer supported as docs are automatically persisted. NzJYou must specify a persist_directory oncreation to persist the collection.rrNr)rWrrRrXrYrZrVpersist)rbrRrcrdr"s r#rzChroma.persists"  " *6  #.44S99ua u::??s5zzA~~ L " " " " " ?~~r% document_iddocumentrc2||g|gS)zUpdate a document in the collection. Args: document_id (str): ID of the document to update. document (Document): Document to update. )update_documents)rbrrs r#update_documentzChroma.update_documents$$k]XJ???r%r1cd|D}d|D}|jtd|j|}t|jjdst|jjdr[ddlm}||jj|||| D]8}|j|d|d |d |d  9dS|j|||| dS)zUpdate a document in the collection. Args: ids (List[str]): List of ids of the document to update. documents (List[Document]): List of documents to update. cg|] }|j Srr-r rs r#r$z+Chroma.update_documents..s@@@(%@@@r%cg|] }|j Srr.rs r#r$z+Chroma.update_documents..s@@@(H%@@@r%Nz?For update, you must specify an embedding function on creation.get_max_batch_sizemax_batch_sizercreate_batches)apirr2r1rir,r/)rrir1r2) r^rrrr`rVchromadb.utils.batch_utilsr update)rbrr1textr.rir batchs r#rzChroma.update_documentssnA@i@@@@@i@@@  # +Q -==dCC    $       $&6    B A A A A A'$,"%    ''a$Qx#Ah#Ah (     # #%" $     r%cls Type[Chroma]c ||d |||||| d| } | d|D}t| jdst| jdr\ddlm} | | j|||D]@} | | d r| d ng| d r| d nd| d An| ||| | S) aCreate a Chroma vectorstore from a raw documents. If a persist_directory is specified, the collection will be persisted there. Otherwise, the data will be ephemeral in-memory. Args: texts (List[str]): List of texts to add to the collection. collection_name (str): Name of the collection to create. persist_directory (Optional[str]): Directory to persist the collection. embedding (Optional[Embeddings]): Embedding function. Defaults to None. metadatas (Optional[List[dict]]): List of metadatas. Defaults to None. ids (Optional[List[str]]): List of document IDs. Defaults to None. client_settings (Optional[chromadb.config.Settings]): Chroma client settings collection_metadata (Optional[Dict]): Collection configurations. Defaults to None. Returns: Chroma: Chroma vectorstore. )r>r?rArCrGrENcNg|]"}ttj#Srrrs r#r$z%Chroma.from_texts..;rr%rrrr)r!rr2r1r"r/)rr2rr)rrVr#r r)r'rrr2rr>rArCrGrErrchroma_collectionr r&s r# from_textszChroma.from_textssPB C +(/+ 3      ;44e444C   %    S  %    S B A A A A A'%-#    "++&+Ah6%((B*/(.us777c!777r%cg|] }|j Srrr/s r#r$z)Chroma.from_documents..vs777cS\777r%) rrr2rr>rArCrGrEr)r,) r'r1rrr>rArCrGrErrrr2s r#from_documentszChroma.from_documentsTspB87Y77777Y777 s~  +/+ 3      r%c .|jjdd|i|dS)zUDelete by vector IDs. Args: ids: List of ids to delete. rNr)r`delete)rbrrrs r#r3z Chroma.deletes+ 22C2622222r%c4|jS)z0Count the number of documents in the collection.)r`countrhs r#__len__zChroma.__len__s%%'''r%)r>r<r?r@rArBrCrDrErFrGrHrIrJrrK)rr@)NNrNN)rjrlrkrmrnrZrorprqrprrrrr)rvr<rr<)NN) rrr2rrrlrrrrr) rrr2rrrlrrrrr) rtr<rrZrrprrrrr) rrrrZrrprqrprrrrr) rrrrZrrprqrprrrrr)) rtr<rrZrrprqrprrrrr))rr) rvr<rrZrrprrrrr) rvr<rrZrrprrrrr))rrrrZrrZrrrrprqrprrrrr)rtr<rrZrrZrrrrprqrprrrrr)rrK)NNNNNN)rrrorrr r r rqr rrlrr )rr<rrrrK)rrr1rrrK)r'r(rrrr@r2rrrlr>r<rArBrCrDrGrHrErFrrrrr:)r'r(r1rrr@rrlr>r<rArBrCrDrGrHrErFrrrrr:rg)rrlrrrrrK)rrZ)#__name__ __module__ __qualname____doc__r=__annotations__repropertyrirrrrr DEFAULT_Krrrrrrrrrrrr rrr classmethodr,r1r3r6rr%r#r:r:5s  /:&9999 B37+/>B.2,0AE>>>>>>>>>>@(((X(X122,08<*.37     32 4GGGG+/#' QQQQQl+/#' MMMMMd+/ 333332+/37 )))))8+/37 44444B+/37 &4&4&4&4&4PF+/ )))))\+/ *****^ +/37. . . . . f +/37)))))V>>>> (,!%# $26'+$.$.$.$.$.LZ = ####*@@@@****X+/*.#'A+/>B,0.2A!A!A!A![A!F+/#'A+/>B .2- - - - [- ^33333((((((r%r:)rrrr)rrrr))* __future__rrzloggingrtypingrrrrrr r r r numpyrlangchain_core._apir langchain_core.documentsrlangchain_core.embeddingsrlangchain_core.utilsrlangchain_core.vectorstoresr&langchain_community.vectorstores.utilsrrRrSchromadb.api.typesrrrr getLoggerloggerr=r(r&r:rr%r#rLs""""""                        ******------000000))))))333333MMMMMMGOOOFFFFFFFFFFFF     DDDD     '5=VWWWX (X (X (X (X ([X (X (XWX (X (X (r%