文档上传接口支持ragflow上传并解析

app/api/files.py

@@ -1,4 +1,4 @@
-from fastapi import Depends, APIRouter, HTTPException, UploadFile, File, requests
+from fastapi import Depends, APIRouter, HTTPException, UploadFile, File, requests, Query
 from sqlalchemy.orm import Session
 
 from app.api import Response, get_current_user, ResponseList
@@ -16,6 +16,7 @@ router = APIRouter()
 @router.post("/upload/{agent_id}", response_model=Response)
 async def upload_file(agent_id: str,
                       file: UploadFile = File(...),
+                      chat_id: str = Query(..., description="The ID of the chat"),
                       db: Session = Depends(get_db),
                       current_user: UserModel = Depends(get_current_user)
                       ):
@@ -29,7 +30,18 @@ async def upload_file(agent_id: str,
         return Response(code=400, msg=str(e))
 
     if agent.agent_type == AgentType.RAGFLOW:
-        pass
+        token = get_ragflow_token(db, current_user.id)
+        ragflow_service = RagflowService(base_url=settings.ragflow_base_url)
+        # 查询会话是否存在，不存在先创建会话
+        history = await ragflow_service.get_session_history(token, chat_id)
+        if len(history) == 0:
+            message = {"role": "user", "message": file.filename}
+            await ragflow_service.set_session(token, agent_id, message, chat_id, True)
+
+        ragflow_service = RagflowService(base_url=settings.ragflow_base_url)
+        token = get_ragflow_token(db, current_user.id)
+        doc_ids = await ragflow_service.upload_and_parse(token, chat_id, file.filename, file_content)
+        return Response(code=200, msg="", data={"doc_ids": doc_ids, "file_name": file.filename})
 
     elif agent.agent_type == AgentType.BISHENG:
         bisheng_service = BishengService(base_url=settings.bisheng_base_url)
@@ -38,6 +50,7 @@ async def upload_file(agent_id: str,
             result = await bisheng_service.upload(token, file.filename, file_content)
         except Exception as e:
             raise HTTPException(status_code=500, detail=str(e))
+        result["file_name"] = file.filename
         return Response(code=200, msg="", data=result)
 
     else:

app/service/ragflow.py

@@ -1,5 +1,9 @@
 import httpx
 from typing import Union, Dict, List
+
+from fastapi import HTTPException
+from starlette import status
+
 from app.config.config import settings
 from app.utils.rsa_crypto import RagflowCrypto
 
@@ -8,12 +12,17 @@ class RagflowService:
     def __init__(self, base_url: str):
         self.base_url = base_url
 
-    async def _handle_response(self, response: httpx.Response) -> Union[Dict, List]:
+    def _handle_response(self, response: httpx.Response) -> Union[Dict, List]:
         if response.status_code != 200:
             return {}
 
         data = response.json()
         ret_code = data.get("retcode")
+        if ret_code == 401:
+            raise HTTPException(
+                status_code=status.HTTP_401_UNAUTHORIZED,
+                detail="登录过期",
+            )
         if ret_code != 0:
             return {}
 
@@ -80,7 +89,7 @@ class RagflowService:
         headers = {"Authorization": token}
         async with httpx.AsyncClient() as client:
             response = await client.get(url, headers=headers)
-            data = await self._handle_response(response)
+            data = self._handle_response(response)
             result = [
                 {
                     "id": item["id"],
@@ -91,28 +100,24 @@ class RagflowService:
             ]
             return result
 
-    async def set_session(self, token: str, dialog_id: str, name: str, chat_id: str, is_new: bool) -> list:
+    async def set_session(self, token: str, dialog_id: str, message: dict, chat_id: str, is_new: bool) -> list:
         url = f"{self.base_url}/v1/conversation/set?dialog_id={dialog_id}"
         headers = {"Authorization": token}
         data = {
             "dialog_id": dialog_id,
-            "name": name,
+            "name": message["message"],
             "is_new": is_new,
             "conversation_id": chat_id,
         }
         async with httpx.AsyncClient() as client:
             response = await client.post(url, headers=headers, json=data)
-            data = await self._handle_response(response)
+            data = self._handle_response(response)
             return [
                 {
                     "content": "你好！ 我是你的助理，有什么可以帮到你的吗？",
                     "role": "assistant"
                 },
-                {
-                    "content": name,
-                    "doc_ids": [],
-                    "role": "user"
-                }
+                message,
             ] if data else []
 
     async def get_session_history(self, token: str, chat_id: str) -> list:
@@ -120,5 +125,17 @@ class RagflowService:
         headers = {"Authorization": token}
         async with httpx.AsyncClient() as client:
             response = await client.get(url, headers=headers)
-            data = await self._handle_response(response)
-            return data.get("message", [])
+            data = self._handle_response(response)
+            return data.get("message", [])
+
+    async def upload_and_parse(self, token: str, chat_id: str, filename: str, file: bytes) -> str:
+        url = f"{self.base_url}/v1/document/upload_and_parse"
+        headers = {"Authorization": token}
+        data = {"conversation_id": chat_id}
+
+        # 创建表单数据，包含文件
+        files = {"file": (filename, file)}
+        async with httpx.AsyncClient(timeout=60) as client:
+            response = await client.post(url, headers=headers, files=files, data=data)
+            data = self._handle_response(response)
+            return data