o Jjg(-@s(dZddlmZmZmZmZmZddlmZddl m Z ddl m Z ddl mZddlmZmZddlmZmZdd lmZmZdd lmZmZmZdd lmZmZmZm Z m!Z!dd l"m#Z#d d eeddededeedeede$de$deee$efeffddZ%eddddGddde Z&d S)z7Chain that combines documents by stuffing into context.)AnyDictListOptionalTuple) deprecated) Callbacks)Document)LanguageModelLike)BaseOutputParserStrOutputParser)BasePromptTemplateformat_document)RunnableRunnablePassthrough) ConfigDictFieldmodel_validator)DEFAULT_DOCUMENT_PROMPTDEFAULT_DOCUMENT_SEPARATOR DOCUMENTS_KEYBaseCombineDocumentsChain_validate_prompt)LLMChainN) output_parserdocument_promptdocument_separatordocument_variable_namellmpromptrrrrreturncsft||pt|p t}dtdtffdd }tjd i|ijdd|B|B|BjddS) aCreate a chain for passing a list of Documents to a model. Args: llm: Language model. prompt: Prompt template. Must contain input variable "context" (override by setting document_variable), which will be used for passing in the formatted documents. output_parser: Output parser. Defaults to StrOutputParser. document_prompt: Prompt used for formatting each document into a string. Input variables can be "page_content" or any metadata keys that are in all documents. "page_content" will automatically retrieve the `Document.page_content`, and all other inputs variables will be automatically retrieved from the `Document.metadata` dictionary. Default to a prompt that only contains `Document.page_content`. document_separator: String separator to use between formatted document strings. document_variable_name: Variable name to use for the formatted documents in the prompt. Defaults to "context". Returns: An LCEL Runnable. The input is a dictionary that must have a "context" key that maps to a List[Document], and any other input variables expected in the prompt. The Runnable return type depends on output_parser used. Example: .. code-block:: python # pip install -U langchain langchain-community from langchain_community.chat_models import ChatOpenAI from langchain_core.documents import Document from langchain_core.prompts import ChatPromptTemplate from langchain.chains.combine_documents import create_stuff_documents_chain prompt = ChatPromptTemplate.from_messages( [("system", "What are everyone's favorite colors:\n\n{context}")] ) llm = ChatOpenAI(model="gpt-3.5-turbo") chain = create_stuff_documents_chain(llm, prompt) docs = [ Document(page_content="Jesse loves red but not yellow"), Document(page_content = "Jamal loves green but not as much as he loves orange") ] chain.invoke({"context": docs}) inputsr csfdd|DS)Nc3s|]}t|VqdSN)r.0doc)_document_prompt`/var/www/html/zoom/venv/lib/python3.10/site-packages/langchain/chains/combine_documents/stuff.py Ts  zDcreate_stuff_documents_chain..format_docs..)join)r!r&rrr'r( format_docsSsz1create_stuff_documents_chain..format_docs format_inputs)run_namestuff_documents_chainNr')rrr dictstrrassign with_config)rrrrrr_output_parserr,r'r+r(create_stuff_documents_chains 7 r5z0.2.13z1.0zThis class is deprecated. Use the `create_stuff_documents_chain` constructor instead. See migration guide here: https://python.langchain.com/v0.2/docs/versions/migrating_chains/stuff_docs_chain/)sinceremovalmessagec s4eZdZUdZeed< edddZeed< e ed< dZ e ed < e d d d Z e d dededefddZedee ffdd ZdeededefddZdeededeefddZ d#deedededee effddZ d#deedededee effdd Zede fd!d"ZZ S)$StuffDocumentsChaina-Chain that combines documents by stuffing into context. This chain takes a list of documents and first combines them into a single string. It does this by formatting each document into a string with the `document_prompt` and then joining them together with `document_separator`. It then adds that new string to the inputs with the variable name set by `document_variable_name`. Those inputs are then passed to the `llm_chain`. Example: .. code-block:: python from langchain.chains import StuffDocumentsChain, LLMChain from langchain_core.prompts import PromptTemplate from langchain_community.llms import OpenAI # This controls how each document will be formatted. Specifically, # it will be passed to `format_document` - see that function for more # details. document_prompt = PromptTemplate( input_variables=["page_content"], template="{page_content}" ) document_variable_name = "context" llm = OpenAI() # The prompt here should take as an input variable the # `document_variable_name` prompt = PromptTemplate.from_template( "Summarize this content: {context}" ) llm_chain = LLMChain(llm=llm, prompt=prompt) chain = StuffDocumentsChain( llm_chain=llm_chain, document_prompt=document_prompt, document_variable_name=document_variable_name ) llm_chaincCstSr")rr'r'r'r(szStuffDocumentsChain.)default_factoryrrz rTforbid)arbitrary_types_allowedextrabefore)modevaluesr cCs`|djj}d|vrt|dkr|d|d<|Std|d|vr.td|dd||S)zGet default document variable name, if not provided. If only one variable is present in the llm_chain.prompt, we can infer that the formatted documents should be passed in with this variable name. r:rrzQdocument_variable_name must be provided if there are multiple llm_chain_variableszdocument_variable_name z- was not found in llm_chain input_variables: )rinput_variableslen ValueError)clsrBllm_chain_variablesr'r'r("get_default_document_variable_names      z6StuffDocumentsChain.get_default_document_variable_namecs"fddjjD}tj|S)Ncsg|] }|jkr|qSr')r)r$kselfr'r( sz2StuffDocumentsChain.input_keys..)r: input_keyssuper)rL extra_keys __class__rKr(rNs  zStuffDocumentsChain.input_keysdocskwargsc s>fdd|D}fdd|D}j||j<|S)aConstruct inputs from kwargs and docs. Format and then join all the documents together into one input with name `self.document_variable_name`. Also pluck any additional variables from **kwargs. Args: docs: List of documents to format and then join into single input **kwargs: additional inputs to chain, will pluck any other required arguments from here. Returns: dictionary of inputs to LLMChain csg|]}t|jqSr')rrr#rKr'r(rMsz3StuffDocumentsChain._get_inputs..cs$i|]\}}|jjjvr||qSr')r:rrD)r$rJvrKr'r( s z3StuffDocumentsChain._get_inputs..)itemsrr*r)rLrSrT doc_stringsr!r'rKr( _get_inputss  zStuffDocumentsChain._get_inputscKs2|j|fi|}|jjjdi|}|j|S)aVReturn the prompt length given the documents passed in. This can be used by a caller to determine whether passing in a list of documents would exceed a certain prompt length. This useful when trying to ensure that the size of a prompt remains below a certain context limit. Args: docs: List[Document], a list of documents to use to calculate the total prompt length. Returns: Returns None if the method does not depend on the prompt length, otherwise the length of the prompt in tokens. Nr')rYr:rformat_get_num_tokens)rLrSrTr!rr'r'r( prompt_lengths z!StuffDocumentsChain.prompt_lengthN callbackscKs,|j|fi|}|jjdd|i|ifS)aStuff all documents into one prompt and pass to LLM. Args: docs: List of documents to join together into one variable callbacks: Optional callbacks to pass along **kwargs: additional parameters to use to get inputs to LLMChain. Returns: The first element returned is the single string output. The second element returned is a dictionary of other keys to return. r]Nr')rYr:predictrLrSr]rTr!r'r'r( combine_docssz StuffDocumentsChain.combine_docscs4|j|fi|}|jjdd|i|IdHifS)aAsync stuff all documents into one prompt and pass to LLM. Args: docs: List of documents to join together into one variable callbacks: Optional callbacks to pass along **kwargs: additional parameters to use to get inputs to LLMChain. Returns: The first element returned is the single string output. The second element returned is a dictionary of other keys to return. r]Nr')rYr:apredictr_r'r'r( acombine_docss z!StuffDocumentsChain.acombine_docscCsdS)Nr/r'rKr'r'r( _chain_typeszStuffDocumentsChain._chain_typer")!__name__ __module__ __qualname____doc__r__annotations__rrr r1rr model_configr classmethodrrrIpropertyrrNr r0rYrintr\rrr`rbrc __classcell__r'r'rQr(r9csZ  %     r9)'rgtypingrrrrrlangchain_core._apirlangchain_core.callbacksrlangchain_core.documentsr langchain_core.language_modelsr langchain_core.output_parsersr r langchain_core.promptsr rlangchain_core.runnablesrrpydanticrrr'langchain.chains.combine_documents.baserrrrrlangchain.chains.llmrr1r5r9r'r'r'r(sJ      K