Q i ddlZddlmZmZmZmZmZddlmZddl m Z ddl m Z m Z mZddlmZmZmZmZmZmZmZmZddlmZmZddlmZdd lmZdd lm Z dd l!m"Z"dd l#m$Z$m%Z%m&Z&m'Z'dd l(m)Z)ddl*m+Z+e"dZ,Gdde Z-dS)N)AnyDictListOptionalTuple)Redis) BaseLLMCache) CacheEntryCacheHitSemanticCacheIndexSchema)CACHE_VECTOR_FIELD_NAMEENTRY_ID_FIELD_NAMEINSERTED_AT_FIELD_NAMEMETADATA_FIELD_NAMEPROMPT_FIELD_NAMEREDIS_KEY_FIELD_NAMERESPONSE_FIELD_NAMEUPDATED_AT_FIELD_NAME)AsyncSearchIndex SearchIndex)VectorRangeQuery)FilterExpressionhashify) get_logger)current_timestampdeprecated_argument serializevalidate_vector_dims)BaseVectorizer)HFTextVectorizerz [RedisVL]c.eZdZUdZeed<dZeeed<e ddddddddd id f d e d e d ee dee deeee efdeede dee efdeffd Zde fdZ d?dedeeee efdefdZdefdZedefdZedeefdZede fdZd e ddfdZd@dZd@dZ dAdeee d eee ddfd!Z dAdeee d eee ddfd"Z d#ee dee fd$Z!d#ee dee fd%Z"d&ee fd'Z# dBd#ee d&eee d)e d*eee d+ee$d ee deee effd,Z% dBd#ee d&eee d)e d*eee d+ee$d ee deee effd-Z& d?d.eee efd*eee de'ee eee efffd/Z( dCd#e d0e d&eee d1eee efd2eee efd ee de fd3Z) dCd#e d0e d&eee d1eee efd2eee efd ee de fd4Z*d5e ddfd6Z+d5e ddfd7Z,d8Z-d9Z.d:Z/d;Z0fd<Z1fd=Z2 d?d#e d2eee efde fd>Z3xZ4S)D SemanticCachez)Semantic Cache for Large Language Models._indexN_aindexdtype vectorizerllmcacheg?zredis://localhost:6379Fnamedistance_thresholdttlfilterable_fields redis_client redis_urlconnection_kwargs overwritec  t|||||| dd} |rSt|tst d| r%|j| krtd| d|j||_n-| } | r| | tddd i| |_| |ttttt t"g|_t'j|||jj|jj} || |} t/d| |j|jd d |jd |_d|_| |_|js|jr|st<d t/j ||jj}|j!"|jj!"krtd|d|j#|jddS)aeSemantic Cache for Large Language Models. Args: name (str, optional): The name of the semantic cache search index. Defaults to "llmcache". distance_threshold (float, optional): Semantic distance threshold for the cache in Redis COSINE units [0-2], where lower values indicate stricter matching. Defaults to 0.1. ttl (Optional[int], optional): The time-to-live for records cached in Redis. Defaults to None. vectorizer (Optional[BaseVectorizer], optional): The vectorizer for the cache. Defaults to HFTextVectorizer. filterable_fields (Optional[List[Dict[str, Any]]]): An optional list of RedisVL fields that can be used to customize cache retrieval with filters. redis_client(Optional[Redis], optional): A redis client connection instance. Defaults to None. redis_url (str, optional): The redis url. Defaults to redis://localhost:6379. connection_kwargs (Dict[str, Any]): The connection arguments for the redis client. Defaults to empty {}. overwrite (bool): Whether or not to force overwrite the schema for the semantic cache index. Defaults to false. Raises: TypeError: If an invalid vectorizer is provided. TypeError: If the TTL value is not an int. ValueError: If the threshold is not between 0 and 2 (Redis COSINE distance). ValueError: If existing schema does not match new schema and overwrite is False. )r)r+r-r.r/r&Nz.Must provide a valid redisvl.vectorizer class.zProvided dtype z! does not match vectorizer dtype )r&modelzredis/langcache-embed-v1r.schemar-r.r/aThe default vectorizer has changed from `sentence-transformers/all-mpnet-base-v2` to `redis/langcache-embed-v1` in version 0.6.0 of RedisVL. For more information about this model, please refer to https://arxiv.org/abs/2504.02268 or visit https://huggingface.co/redis/langcache-embed-v1. To continue using the old vectorizer, please specify it explicitly in the constructor as: vectorizer=HFTextVectorizer(model='sentence-transformers/all-mpnet-base-v2'))r-zExisting index z schema does not match the user provided schema for the semantic cache. If you wish to overwrite the index schema, set overwrite=True during initialization.F)r0drop)$super__init__pop isinstancer TypeErrorr& ValueError _vectorizerupdater! set_thresholdrrrrrr return_fieldsr from_paramsdims_modify_schemar _redis_client redis_kwargsr$r%r0existsloggerwarning from_existingr4to_dictcreate)selfr)r*r+r'r,r-r.r/r0kwargsr&vectorizer_kwargsr4existing_index __class__s C:\Users\Dell Inspiron 16\Desktop\tws\AgrotaPowerBi\back-agrota-powerbi\mcp-client-agrota\venv\Lib\site-packages\redisvl/extensions/cache/llm/semantic.pyr8zSemanticCache.__init__-sV %/     7D))  j.99 R PQQQ )U22 `e``jN^`` *D  !'  6!((u(555/  0 #  D  -...   " !   *5 $(-t/?/E  $$V->??" +' 4   34     #~ $+"4"4"6"6  c)64;#<N$,,..$+2D2L2L2N2NNN kdkkk T^%@@@@@returnc8d|jd|jd|jdS)NzSemanticCache(name=z, distance_threshold=z, ttl=))r)r*r+rLs rQ__repr__zSemanticCache.__repr__sF M$) M M"&"9 M MAE M M M rRr4c|ot|jtgz}|D]O}|d}||vrt|d|||j|P|S)zAModify the base cache schema using the provided filterable fieldsNr)z7 is a reserved field name for the semantic cache schema)setr@rr< add_fieldappend)rLr4r,protected_field_names filter_field field_names rQrCzSemanticCache._modify_schemas  ($'(:>R=S(S$T$T ! 1 6 6 )&1 !666$%^^^  ..."))*5555 rRcK|jO|d{V}td|jj||jdd|jd|_|jS)z.Lazily construct the async search index class.Nr.r3r/r6)r%_get_async_redis_clientrr$r4rE)rL async_clients rQ_get_async_indexzSemanticCache._get_async_indexs < !%!=!=!?!???????L+{))+K8#$78 DL |rRc|jS)zoThe underlying SearchIndex for the cache. Returns: SearchIndex: The search index. )r$rVs rQindexzSemanticCache.indexs {rRc|jS)zThe underlying AsyncSearchIndex for the cache. Returns: AsyncSearchIndex: The async search index. )r%rVs rQaindexzSemanticCache.aindexs |rRc|jS)z}The semantic distance threshold for the cache. Returns: float: The semantic distance threshold. )_distance_thresholdrVs rQr*z SemanticCache.distance_thresholds ''rRcdt|cxkrdksntd|t||_dS)aSets the semantic distance threshold for the cache. Args: distance_threshold (float): The semantic distance threshold for the cache. Raises: ValueError: If the threshold is not between 0 and 2 (Redis COSINE distance). rz&Distance must be between 0 and 2, got N)floatr<rh)rLr*s rQr?zSemanticCache.set_thresholdscE,--22222222M9KMM $));#<#<   rRc<|jddS)z(Delete the cache and its index entirely.Tr5N)r$deleterVs rQrnzSemanticCache.deletes! %%%%%rRcvK|d{V}|dd{VdS)z.Async delete the cache and its index entirely.NTrm)rbrn)rLrfs rQadeletezSemanticCache.adelete sU,,........mmm&&&&&&&&&&&rRidskeysc||td|&jfd|D|j|dSdS)a0Drop specific entries from the cache by ID or Redis key. Args: ids (Optional[List[str]]): List of entry IDs to remove from the cache. Entry IDs are the unique identifiers without the cache prefix. keys (Optional[List[str]]): List of full Redis keys to remove from the cache. Keys are the complete Redis keys including the cache prefix. Note: At least one of ids or keys must be provided. Raises: ValueError: If neither ids nor keys is provided. N-At least one of ids or keys must be provided.cDg|]}j|Sr6r$key.0idrLs rQ z&SemanticCache.drop..$s'"E"E"E24;??2#6#6"E"E"ErR)r<r$ drop_keys)rLrqrrs` rQr5zSemanticCache.dropsy" ;4<LMM M ? K ! !"E"E"E"E"E"E"E F F F   K ! !$ ' ' ' ' '  rRcKd{V}||td|'|fd|Dd{V|||d{VdSdS)a6Async drop specific entries from the cache by ID or Redis key. Args: ids (Optional[List[str]]): List of entry IDs to remove from the cache. Entry IDs are the unique identifiers without the cache prefix. keys (Optional[List[str]]): List of full Redis keys to remove from the cache. Keys are the complete Redis keys including the cache prefix. Note: At least one of ids or keys must be provided. Raises: ValueError: If neither ids nor keys is provided. NrtcDg|]}j|Sr6rvrxs rQr{z'SemanticCache.adrop..@s'#F#F#FBDKOOB$7$7#F#F#FrR)rbr<r|)rLrqrrrfs` rQadropzSemanticCache.adrop(s",,........ ;4<LMM M ?""#F#F#F#F##F#F#FGG G G G G G G G  ""4(( ( ( ( ( ( ( ( ( (  rRpromptct|tstd|j|}|S)\Converts a text prompt to its vector representation using the configured vectorizer.Prompt must be a string.)r:strr;r=embedrLrresults rQ_vectorize_promptzSemanticCache._vectorize_promptDs@&#&& 8677 7!''// rRcKt|tstd|j|d{V}|S)rrN)r:rr;r=aembedrs rQ_avectorize_promptz SemanticCache._avectorize_promptMsV&#&& 8677 7'..v66666666 rRvectorc|jjjtjj}t t||dS)zzChecks the size of the provided vector and raises an error if it doesn't match the search index vector dimensions.N)r$r4fieldsr attrsrBrlen)rLrschema_vector_dimss rQ_check_vector_dimsz SemanticCache._check_vector_dimsVs@"[/6 #   S[[*<=====rR num_resultsr@filter_expressionc t||gstd|r$t|tst d|p|j}||||}|||ntdt|t|j ||d||j j }|j |}|||\} } | D]} || | S)aChecks the semantic cache for results similar to the specified prompt or vector. This method searches the cache using vector similarity with either a raw text prompt (converted to a vector) or a provided vector as input. It checks for semantically similar prompts and fetches the cached LLM responses. Args: prompt (Optional[str], optional): The text prompt to search for in the cache. vector (Optional[List[float]], optional): The vector representation of the prompt to search for in the cache. num_results (int, optional): The number of cached results to return. Defaults to 1. return_fields (Optional[List[str]], optional): The fields to include in each returned result. If None, defaults to all available fields in the cached entry. filter_expression (Optional[FilterExpression]) : Optional filter expression that can be used to filter cache results. Defaults to None and the full cache will be searched. distance_threshold (Optional[float]): The threshold for semantic vector distance. Returns: List[Dict[str, Any]]: A list of dicts containing the requested return fields for each similar cached response. Raises: ValueError: If neither a `prompt` nor a `vector` is specified. ValueError: if 'vector' has incorrect dimensions. TypeError: If `return_fields` is not a list when provided. .. code-block:: python response = cache.check( prompt="What is the capital city of France?" ) *Either prompt or vector must be specified.'Return fields must be a list of values.N0Failed to generate a valid vector for the query.Trvector_field_namer@r*r return_scorerr&)anyr<r:listr;rhrrrr r@r=r&r$query_process_cache_resultsexpire) rLrrrr@rr*rcache_search_results redis_keys cache_hitsrws rQcheckzSemanticCache.check^sL`FF#$$ KIJJ J  GM4!@!@ GEFF F0K43K >f0++F33F    # #F + + + +OPP P!5,1#/"(     ${0077!%!.s%GGGSt||C00GGGrR)rbrr<r:rr;rhrrrr r@r=r&rrasynciogather) rLrrrr@rr*rfrrrrs ` rQacheckzSemanticCache.achecks`,,........FF#$$ KIJJ J  GM4!@!@ GEFF F0K43K >f0226::::::::F    # #F + + + +OPP P!5,1#/"(    &,\\%%8%8888888!%!z8SemanticCache._process_cache_results..)s0"""!QqM?Q?QAq?Q?Q?QrRr6)r9r[r rJr:ritemsr) rLrr@rrcache_search_result redis_key cache_hitcache_hit_dicts ` rQrz$SemanticCache._process_cache_resultss !# +- #7 . . +//55I   i ( ( (!77#677I&..00N-.. = """"%3%9%9%;%;""" 4=N/ 0   n - - - -:%%rRresponsemetadatafilterscR|p||}|||||}t||||||}|p|j}|j||jj g|t} | dS)aMStores the specified key-value pair in the cache along with metadata. Args: prompt (str): The user prompt to cache. response (str): The LLM response to cache. vector (Optional[List[float]], optional): The prompt vector to cache. Defaults to None, and the prompt vector is generated on demand. metadata (Optional[Dict[str, Any]], optional): The optional metadata to cache alongside the prompt and response. Defaults to None. filters (Optional[Dict[str, Any]]): The optional tag to assign to the cache entry. Defaults to None. ttl (Optional[int]): The optional TTL override to use on this individual cache entry. Defaults to the global TTL setting. Returns: str: The Redis key for the entries added to the semantic cache. Raises: ValueError: If neither prompt nor vector is specified. ValueError: if vector has incorrect dimensions. TypeError: If provided metadata is not a dictionary. .. code-block:: python key = cache.store( prompt="What is the capital city of France?", response="Paris", metadata={"city": "Paris", "country": "France"} ) entry_idrr prompt_vectorrrdatar+id_fieldr) rr_make_entry_idr _ttlr$loadrJr=r&r) rLrrrrrr+r cache_entryrrs rQstorezSemanticCache.store3sR9411&99 '''&&vw77!     TY{%%d&6&<==>(   AwrRcK|d{V}|p||d{V}|||||}t ||||||} |p|j}|| |jj g|td{V} | dS)aZAsync stores the specified key-value pair in the cache along with metadata. Args: prompt (str): The user prompt to cache. response (str): The LLM response to cache. vector (Optional[List[float]], optional): The prompt vector to cache. Defaults to None, and the prompt vector is generated on demand. metadata (Optional[Dict[str, Any]], optional): The optional metadata to cache alongside the prompt and response. Defaults to None. filters (Optional[Dict[str, Any]]): The optional tag to assign to the cache entry. Defaults to None. ttl (Optional[int]): The optional TTL override to use on this individual cache entry. Defaults to the global TTL setting. Returns: str: The Redis key for the entries added to the semantic cache. Raises: ValueError: If neither prompt nor vector is specified. ValueError: if vector has incorrect dimensions. TypeError: If provided metadata is not a dictionary. .. code-block:: python key = await cache.astore( prompt="What is the capital city of France?", response="Paris", metadata={"city": "Paris", "country": "France"} ) Nrrr) rbrrrr rrrJr=r&r) rLrrrrrr+rfrrrrs rQastorezSemanticCache.astorewsP,,........@!8!8!@!@@@@@@@ '''&&vw77!     TY[[%%d&6&<==>(!        AwrRrwc |r|D]\}}|t|jjjt gzvrt |d|t kr7t|trt|||<std| tti|}|||||dS)a9Update specific fields within an existing cache entry. If no fields are passed, then only the document TTL is refreshed. Args: key (str): the key of the document to update using kwargs. Raises: ValueError if an incorrect mapping is provided as a kwarg. TypeError if metadata is provided and not of type dict. .. code-block:: python key = cache.store('this is a prompt', 'this is a response') cache.update(key, metadata={"hit_count": 1, "model_name": "Llama-2-7b"}) , is not a valid field within the cache entry3If specified, cached metadata must be a dictionary.mappingN)rrYr$r4 field_namesrr<r:dictrr;r>rr_get_redis_clienthsetrrLrwrMrrclients rQr>zSemanticCache.updates  -   1C 2 >BUAV VWWWW$%W%W%WXXX+++!!T**$-aLLq 'Q , MM02C2E2EF G G G++--F KKVK , , , CrRc0K|r|D]\}}|t|jjjt gzvrt |d|t kr7t|trt|||<std| tti|d{V}|||d{V||d{VdS)a{Async update specific fields within an existing cache entry. If no fields are passed, then only the document TTL is refreshed. Args: key (str): the key of the document to update using kwargs. Raises: ValueError if an incorrect mapping is provided as a kwarg. TypeError if metadata is provided and not of type dict. .. code-block:: python key = await cache.astore('this is a prompt', 'this is a response') await cache.aupdate( key, metadata={"hit_count": 1, "model_name": "Llama-2-7b"} ) rrNr)rrYr$r4rrr<r:rrr;r>rrr`rrrs rQaupdatezSemanticCache.aupdatesK&  3   1C 2 >BUAV VWWWW$%W%W%WXXX+++!!T**$-aLLq 'Q , MM02C2E2EF G G G 7799999999F++c6+22 2 2 2 2 2 2 2ll3rRc|S)zContext manager entry point.r6rVs rQ __enter__zSemanticCache.__enter__s rRc.|dS)zContext manager exit point.N) disconnectrLexc_typeexc_valexc_tbs rQ__exit__zSemanticCache.__exit__s rRc K|S)zAsync context manager entry.r6rVs rQ __aenter__zSemanticCache.__aenter__s  rRc>K|d{VdS)zAsync context manager exit.N) adisconnectrs rQ __aexit__zSemanticCache.__aexit__s0           rRct|dr |jr|jt|dr |jr|jt dS)zmDisconnect from Redis and search index. Closes all Redis connections and index connections. r$r%N)hasattrr$rr%disconnect_syncr7rLrPs rQrzSemanticCache.disconnect#s 4 " " %t{ % K " " $ $ $ 4 # # +  + L ( ( * * * rRcKt|dr-|jr&|jd{Vd|_td{VdS)z|Asynchronously disconnect from Redis and search index. Closes all Redis connections and index connections. r%N)rr%rr7rrs rQrzSemanticCache.adisconnect3s 4 # #   ,))++ + + + + + + +DLgg!!###########rRc"t||S)a8Generate a deterministic entry ID for the given prompt and optional filters. Args: prompt (str): The prompt text. filters (Optional[Dict[str, Any]]): Optional filter dictionary. Returns: str: A deterministic entry ID based on the prompt and filters. r)rLrrs rQrzSemanticCache._make_entry_id@svw'''rR)N)rSN)NN)NNrNNN)NNNN)5__name__ __module__ __qualname____doc__r__annotations__r%rrrrrkintr rrrrboolr8rWr rCrbpropertyrdrfr*r?rnrpr5rrrrrrrrrrrr>rrrrrrrr __classcell__)rPs@rQr#r#'s33 *.GX& '...,//$'!/3<@(,1,.AAAAAA"AAc] AA ^, AA $Dc3h$89 AAuoAAAA S>AAAAAAAAAAAA0/AAF #    =A($Dc3h$89 " , (8    {X!12X(E(((X(==$==== &&&&'''' LP((DI&(5=d3i5H( ((((6LP))DI&)5=d3i5H) ))))8 $u+x}e>e>>>>!%(,-18<.2YY Ye%Y Y  S * Y $$45 Y%UOY d38n YYYYz!%(,-18<.2ZZ Ze%Z Z  S * Z $$45 Z%UOZ d38n ZZZZ~.2&&"4S>2& S *& tCy$tCH~.. / &&&&D)--1,0!BBBBe% B 4S>* B $sCx.) Bc]B BBBBP)--1,0!DDDDe% D 4S>* D $sCx.) Dc]D DDDDL'#'D''''R+ + 4+ + + + Z!!! $ $ $ $ $@D ( ( ($,T#s(^$< (  ( ( ( ( ( ( ( (rRr#).rtypingrrrrrredisr!redisvl.extensions.cache.llm.baser #redisvl.extensions.cache.llm.schemar r r redisvl.extensions.constantsr rrrrrrr redisvl.indexrr redisvl.queryrredisvl.query.filterrredisvl.redis.utilsrredisvl.utils.logrredisvl.utils.utilsrrrrredisvl.utils.vectorize.baser (redisvl.utils.vectorize.text.huggingfacer!rGr#r6rRrQrs33333333333333::::::                    87777777******111111''''''(((((( 877777EEEEEE K e (e (e (e (e (Le (e (e (e (e (rR