[docs]classMongoDBStructuredQueryTranslator(Visitor):"""Translator between MongoDB Query API and LangChain's StructuredQuery. With Vector Search Indexes, one can index boolean, date, number, objectId, string, and UUID fields to pre-filter your data. Filtering your data is useful to narrow the scope of your semantic search and ensure that not all vectors are considered for comparison. It reduces the number of documents against which to run similarity comparisons, which can decrease query latency and increase the accuracy of search results. """"""Available logical comparators."""allowed_comparators=[Comparator.EQ,Comparator.NE,Comparator.GT,Comparator.GTE,Comparator.LT,Comparator.LTE,Comparator.IN,Comparator.NIN,]"""Available logical operators."""allowed_operators=[Operator.AND,Operator.OR]## Convert an operator or a comparator to Mongo Query Formatdef_format_func(self,func:Union[Operator,Comparator])->str:self._validate_func(func)map_dict={Operator.AND:"$and",Operator.OR:"$or",Comparator.EQ:"$eq",Comparator.NE:"$ne",Comparator.GTE:"$gte",Comparator.LTE:"$lte",Comparator.LT:"$lt",Comparator.GT:"$gt",Comparator.IN:"$in",Comparator.NIN:"$nin",}returnmap_dict[func]
[docs]classMongoDBAtlasSelfQueryRetriever(SelfQueryRetriever):"""Retriever that uses an LLM to deduce filters for Vector Search algorithm. This can greatly increase power of vector search on collections with structured metadata. Before calling the search algorithm of the vector store, this retriever first prompts an LLM to find logical statements (e.g. and, in) in a semantic query. From the response, it forms a structured query, which it passes to a VectorStore as filters, The fields to look for conditions are specified by ``metadata_field_info`` a simple list of attribute information for each fieldname, type, description. See `How to do "self-querying" retrieval <https://python.langchain.com/docs/how_to/self_query/>`_ for more information. One must index the fields that you want to filter your data by as the filter type in a vectorSearch type index definition. Example usage: .. code-block:: python from langchain_mongodb.retrievers import MongoDBAtlasSelfQueryRetriever from langchain_mongodb import MongoDBAtlasVectorSearch from langchain_ollama.embeddings import OllamaEmbeddings # Start with the standard MongoDB Atlas vector store vectorstore = MongoDBAtlasVectorSearch.from_connection_string( connection_string="mongodb://127.0.0.1:40947/?directConnection=true", namespace=f"{DB_NAME}.{COLLECTION_NAME}", embedding=OllamaEmbeddings(model="all-minilm:l6-v2") ) # Define metadata describing the data metadata_field_info = [ AttributeInfo( name="genre", description="The genre of the movie. One of ['science fiction', 'comedy', 'drama', 'thriller', 'romance', 'animated']", type="string", ), AttributeInfo( name="year", description="The year the movie was released", type="integer", ), AttributeInfo( name="rating", description="A 1-10 rating for the movie", type="float" ), ] # Create search index with filters vectorstore.create_vector_search_index( dimensions=dimensions, filters=[f.name for f in metadata_field_info], wait_until_complete=TIMEOUT ) # Add documents, including embeddings vectorstore.add_documents(fictitious_movies) # Create the retriever from the VectorStore, an LLM and info about the documents retriever = MongoDBAtlasSelfQueryRetriever.from_llm( llm=llm, vectorstore=vectorstore, metadata_field_info=metadata_field_info, document_contents="Descriptions of movies", enable_limit=True ) # This example results in the following composite filter sent to $vectorSearch: # {'filter': {'$and': [{'year': {'$lt': 1960}}, {'rating': {'$gt': 8}}]}} print(retriever.invoke("Movies made before 1960 that are rated higher than 8")) See Also: * `Run Vector Search Queries <https://www.mongodb.com/docs/atlas/atlas-vector-search/vector-search-stage/#run-vector-search-queries>`_ * `How to Index Fields for Vector Search <https://www.mongodb.com/docs/atlas/atlas-vector-search/vector-search-type>`_ * :class:`~langchain_mongodb.vectorstores.MongoDBAtlasVectorSearch` """vectorstore:MongoDBAtlasVectorSearch"""The underlying vector store from which documents will be retrieved."""structured_query_translator:MongoDBStructuredQueryTranslator"""Translator for turning LangChain internal query language into Atlas search params."""query_constructor:Runnable[dict,StructuredQuery]=Field(alias="llm_chain")"""The query constructor chain for generating the vector store queries."""search_type:str="similarity""""The search type to perform on the vector store."""search_kwargs:dict=Field(default_factory=dict)"""Keyword arguments to pass to MongoDBAtlasVectorSearch (e.g. ``{'k':10})``."""verbose:bool=False"""logs the structured query generated by the LLM"""use_original_query:bool=False"""Use original query instead of the LLM's revised query that removes statements with filters."""
[docs]@classmethoddeffrom_llm(cls,llm:BaseLanguageModel,vectorstore:VectorStore,document_contents:str,metadata_field_info:Sequence[Union[AttributeInfo,dict]],enable_limit:bool=False,use_original_query:bool=False,**kwargs:Any,)->SelfQueryRetriever:"""Create a self-querying retriever from an LLM, vector store, and document metadata. This method does NOT create the vector search index. See example usage. Args: llm: A Reasoning model that will produce the structured query. vectorstore: MongoDBAtlasVectorSearch. document_contents: Description of the data in the collection. metadata_field_info: Fields must be present in vector search index. enable_limit: Whether to instruct the LLM to look for statements involving limits. use_original_query: By default, sentences defining filters and limits are removed from query. Set to True if you wish to include these in the text to embed. **kwargs: Additional arguments to pass to retriever constructor (e.g. search_kwargs) Returns: A retriever invoked by a text query. """returnsuper().from_llm(llm=llm,vectorstore=vectorstore,document_contents=document_contents,metadata_field_info=metadata_field_info,structured_query_translator=MongoDBStructuredQueryTranslator(),enable_limit=enable_limit,use_original_query=use_original_query,**kwargs,)
[docs]defclose(self)->None:"""Close the resources used by the retriever."""self.vectorstore.close()
