rag-gateway/app/api/excel.py

from fastapi import APIRouter, File, UploadFile, Depends
from fastapi.responses import JSONResponse, FileResponse
from fastapi.exceptions import HTTPException
from sqlalchemy.orm import Session
from starlette.websockets import WebSocket, WebSocketDisconnect
from werkzeug.utils import secure_filename
from app.api import get_current_user_websocket
from app.models.agent_model import AgentModel, AgentType
from app.models.base_model import get_db
from app.models.user_model import UserModel
from app.utils.excelmerge.conformity import run_conformity
import shutil
import os

router = APIRouter()

ALLOWED_EXTENSIONS = {'xlsx'}
EXCEL_FILES_PATH = 'data/output'
SOURCE_FILES_PATH = 'data/source'

def allowed_file(filename):
    return '.' in filename and filename.rsplit('.', 1)[1].lower() in ALLOWED_EXTENSIONS


def create_dir_if_not_exists(path):
    if not os.path.exists(path):
        os.makedirs(path)


# 清理函数
def clear_directory(path):
    for filename in os.listdir(path):
        file_path = os.path.join(path, filename)
        try:
            if os.path.isfile(file_path) or os.path.islink(file_path):
                os.unlink(file_path)
            elif os.path.isdir(file_path):
                shutil.rmtree(file_path)
        except Exception as e:
            return {"error": "清空出错"}
    return {"message": "目录已清空"}


@router.post('/excel/upload')
async def upload_file(files: list[UploadFile] = File(...)):
    if not any(file.filename for file in files):
        return JSONResponse(content={"error": "没有文件部分"}, status_code=400)

    create_dir_if_not_exists(SOURCE_FILES_PATH)
    create_dir_if_not_exists(EXCEL_FILES_PATH)
    clear_directory(SOURCE_FILES_PATH)
    clear_directory(EXCEL_FILES_PATH)

    save_path_list = []
    for file in files:
        if file.filename == '':
            return JSONResponse(content={"error": "没有选择文件"}, status_code=400)
        if file and allowed_file(file.filename):
            filename = secure_filename(file.filename)
            save_path = os.path.join(SOURCE_FILES_PATH, filename)
            with open(save_path, 'wb') as buffer:
                shutil.copyfileobj(file.file, buffer)
            save_path_list.append(save_path)
        else:
            return JSONResponse(content={"error": "不允许的文件类型"}, status_code=400)
    return JSONResponse(content={"code": 200, "msg": "", "data": {}}, status_code=200)


# ws://localhost:9201/api/document/ws/excel
@router.websocket("/ws/excel")
async def ws_excel(websocket: WebSocket):
    await websocket.accept()
    while True:
        data = await websocket.receive_text()
        try:
            if data == "合并Excel文件":
                clear_directory(EXCEL_FILES_PATH)
                output_file_path = run_conformity()
                await websocket.send_json({"step_message": "开始合并", "type": "stream", "files": []})
            elif data == "查询合并进度":
                files = os.listdir(EXCEL_FILES_PATH)
                if not files:
                    await websocket.send_json({"step_message": "正在合并中", "type": "stream", "files": []})
                else:
                    await websocket.send_json({"step_message": "文档合并成功！", "type": "stream", "files": []})
            elif data == "获取文件":
                files = os.listdir(EXCEL_FILES_PATH)
                if not files:
                    await websocket.send_json({"error": "目录下没有生成的文件", "type": "stream", "files": []})
                else:
                    first_file = files[0]
                    file_name = os.path.basename(first_file)
                    file_url = f"./api/document/download/{first_file}"
                    await websocket.send_json({
                        "step_message": "文档合并成功！",
                        "type": "stream",
                        "files": [{
                            "file_name": file_name,
                            "file_url": file_url
                        }]
                    })
            else:
                await websocket.send_json({"error": "未知指令"})
        except Exception as e:
            await websocket.send_json({"error": str(e)})


@router.get("/download/{filename}")
async def download_file(filename: str):
    try:
        return FileResponse(os.path.join(EXCEL_FILES_PATH, filename), filename=filename,
                            media_type='application/vnd.openxmlformats-officedocument.spreadsheetml.sheet')
    except FileNotFoundError:
        raise HTTPException(status_code=404, detail="文件不存在")
    except Exception as e:
        raise HTTPException(status_code=500, detail="服务器错误")