智能文档问答系统

┌──────────────────────────────────────────┐
│              Streamlit UI                │
│    (文档上传 / 提问输入 / 回答展示)         │
└──────────────────┬───────────────────────┘
                   │
┌──────────────────▼───────────────────────┐
│              FastAPI Backend             │
│       /upload  /query  /history          │
└───────┬──────────────────┬───────────────┘
        │                  │
┌───────▼───────┐  ┌───────▼───────────┐
│  Document     │  │   RAG Pipeline    │
│  Processing   │  │                   │
│  • PDF Parser  │  │  • Embedding     │
│  • Splitter   │  │  • Vector Search │
│  • Cleaner    │  │  • LLM Generate  │
└───────┬───────┘  └───────────────────┘
        │
┌───────▼───────┐
│   ChromaDB    │
│  (向量存储)    │
└───────────────┘

from langchain.document_loaders import (
    PyPDFLoader,
    Docx2txtLoader,
    TextLoader,
    UnstructuredMarkdownLoader
)

class DocumentProcessor:
    LOADERS = {
        ".pdf": PyPDFLoader,
        ".docx": Docx2txtLoader,
        ".txt": TextLoader,
        ".md": UnstructuredMarkdownLoader,
    }

    @staticmethod
    def load(file_path: str):
        ext = Path(file_path).suffix.lower()
        loader_cls = DocumentProcessor.LOADERS.get(ext)
        if not loader_cls:
            raise ValueError(f"不支持的文件格式: {ext}")
        loader = loader_cls(file_path)
        return loader.load()

from fastapi import FastAPI, UploadFile, HTTPException
from pydantic import BaseModel

app = FastAPI(title="文档问答系统")

class QueryRequest(BaseModel):
    question: str
    collection_name: str
    top_k: int = 4

class QueryResponse(BaseModel):
    answer: str
    sources: list[str]
    confidence: float

@app.post("/api/upload")
async def upload_document(file: UploadFile, collection: str):
    # 文档处理 + 向量化 + 存储
    pass

@app.post("/api/query", response_model=QueryResponse)
async def query_document(request: QueryRequest):
    # RAG 检索 + 生成
    pass

# 1. 安装依赖
pip install -r requirements.txt

# 2. 配置环境变量
export OPENAI_API_KEY=your_key_here

# 3. 启动服务
uvicorn main:app --host 0.0.0.0 --port 8000

# 4. 启动 Streamlit UI
streamlit run app.py