from langchain_community.document_loaders import CSVLoader
from langchain_community.vectorstores import Pinecone
from langchain_openai import OpenAIEmbeddings
from langchain_text_splitters import CharacterTextSplitter

# 加載數(shù)據(jù)
loader = CSVLoader("customer_reviews.csv")
documents = loader.load()

# 文本分割
text_splitter = CharacterTextSplitter(chunk_size=500, chunk_overlap=50)
texts = text_splitter.split_documents(documents)

# 向量化
embeddings = OpenAIEmbeddings()
vectorstore = Pinecone.from_documents(texts, embeddings)

# 配置Retriever
retriever = vectorstore.as_retriever()

docs = retriever.invoke("What do customers think about the battery life?")

from langchain.retrievers.multi_query import MultiQueryRetriever
from langchain_openai import ChatOpenAI

# 創(chuàng)建查詢
question = "What features do customers value in smartphones?"
llm = ChatOpenAI(temperature=0)

# 初始化MultiQueryRetriever
retriever_from_llm = MultiQueryRetriever.from_llm(
    retriever=vectordb.as_retriever(), llm=llm
)
unique_docs = retriever_from_llm.invoke(question)

from langchain.retrievers.multi_query import MultiQueryRetriever
from langchain_openai import ChatOpenAI

# 創(chuàng)建查詢
question = "What features do customers value in smartphones?"
llm = ChatOpenAI(temperature=0)

# 初始化MultiQueryRetriever
retriever_from_llm = MultiQueryRetriever.from_llm(
    retriever=vectordb.as_retriever(), llm=llm
)
unique_docs = retriever_from_llm.invoke(question)

from langchain.retrievers import ContextualCompressionRetriever
from langchain.retrievers.document_compressors import LLMChainExtractor
from langchain_openai import OpenAI

# 初始化LLM和壓縮器
llm = OpenAI(temperature=0)
compressor = LLMChainExtractor.from_llm(llm)

# 配置上下文壓縮檢索器
compression_retriever = ContextualCompressionRetriever(
    base_compressor=compressor, base_retriever=retriever
)
compressed_docs = compression_retriever.invoke("What actions are being proposed to combat climate change?")

from typing import List
from langchain_core.documents import Document
from langchain_core.retrievers import BaseRetriever

class CustomRetriever(BaseRetriever):
    """一個簡單的檢索器，根據(jù)查詢返回包含關(guān)鍵字的文檔。"""

    documents: List[Document]
    k: int

    def _get_relevant_documents(self, query: str) -> List[Document]:
        matching_documents = [doc for doc in self.documents if query.lower() in doc.page_content.lower()]
        return matching_documents[:self.k]

# 示例
documents = [
    Document("Dogs are great companions.", {"type": "dog"}),
    Document("Cats are independent pets.", {"type": "cat"}),
]
retriever = CustomRetriever(documents=documents, k=1)
result = retriever.invoke("dog")
print(result[0].page_content)  # Output: "Dogs are great companions."