68 lines
2.4 KiB
Python
68 lines
2.4 KiB
Python
# rag_core/retriever_factory.py
|
||
from langchain_core.embeddings import Embeddings
|
||
from langchain_classic.retrievers import ParentDocumentRetriever
|
||
from langchain_text_splitters import RecursiveCharacterTextSplitter, TextSplitter
|
||
from langchain_core.stores import BaseStore
|
||
|
||
from rag_core import LlamaCppEmbedder, QdrantVectorStore, create_docstore
|
||
|
||
|
||
def create_parent_retriever(
|
||
collection_name: str = "rag_documents",
|
||
parent_splitter: TextSplitter | None = None,
|
||
child_splitter: TextSplitter | None = None,
|
||
docstore: BaseStore | None = None,
|
||
search_k: int = 5,
|
||
parent_chunk_size: int = 1000,
|
||
parent_chunk_overlap: int = 100,
|
||
child_chunk_size: int = 200,
|
||
child_chunk_overlap: int = 20,
|
||
) -> ParentDocumentRetriever:
|
||
"""
|
||
创建 ParentDocumentRetriever 实例。
|
||
|
||
Args:
|
||
collection_name: Qdrant 集合名称，默认 "rag_documents"
|
||
parent_splitter: 父文档切分器，默认 None（使用默认参数创建）
|
||
child_splitter: 子文档切分器，默认 None（使用默认参数创建）
|
||
docstore: 文档存储实例，默认 None（使用默认参数创建）
|
||
search_k: 检索时返回的结果数，默认 5
|
||
parent_chunk_size: 父文档块大小，默认 1000
|
||
parent_chunk_overlap: 父文档块重叠大小，默认 100
|
||
child_chunk_size: 子文档块大小，默认 200
|
||
child_chunk_overlap: 子文档块重叠大小，默认 20
|
||
|
||
Returns:
|
||
ParentDocumentRetriever 实例
|
||
"""
|
||
# 嵌入模型
|
||
embedder = LlamaCppEmbedder()
|
||
embeddings = embedder.as_langchain_embeddings()
|
||
|
||
# 向量存储（只读）
|
||
vector_store = QdrantVectorStore(collection_name=collection_name)
|
||
|
||
# 切分器（若未提供则创建默认）
|
||
if parent_splitter is None:
|
||
parent_splitter = RecursiveCharacterTextSplitter(
|
||
chunk_size=parent_chunk_size,
|
||
chunk_overlap=parent_chunk_overlap,
|
||
)
|
||
if child_splitter is None:
|
||
child_splitter = RecursiveCharacterTextSplitter(
|
||
chunk_size=child_chunk_size,
|
||
chunk_overlap=child_chunk_overlap,
|
||
)
|
||
|
||
# 文档存储
|
||
if docstore is None:
|
||
docstore, _ = create_docstore()
|
||
|
||
return ParentDocumentRetriever(
|
||
vectorstore=vector_store.get_langchain_vectorstore(),
|
||
docstore=docstore,
|
||
child_splitter=child_splitter,
|
||
parent_splitter=parent_splitter,
|
||
search_kwargs={"k": search_k},
|
||
)
|