添加默认问题回答

api.py

 from collections import Counter
 from contextlib import asynccontextmanager
-from datetime import datetime
 
 import uvicorn
 from fastapi import FastAPI, Request, APIRouter
@@ -13,7 +12,6 @@ from config.system import config
 from model.entity.AiChatPermissionEntity import AiChatPermissionEntity
 from model.entity.AiChatRecommendQuestionEntity import AiChatRecommendQuestionEntity
 from model.entity.AiChatRecordEntity import AiChatRecordEntity
-from model.param.AiChatRecordParam import AiChatRecordParam
 from model.vo.ResultVo import ResultVo
 
 
@@ -96,21 +94,6 @@ async def chat(request: Request, message: str):
                                media_type="text/event-stream", headers={"Cache-Control": "no-cache"})
 
 
-@router.post("/chat/record", description="记录AI回答")
-async def record_chat(chat_record_param: AiChatRecordParam):
-    await db_util.update(
-        AiChatRecordEntity,
-        {
-            "id": chat_record_param.id
-        },
-        {
-            "answer": chat_record_param.answer,
-            "answer_time": datetime.now().strftime("%Y-%m-%d %H:%M:%S")
-        }
-    )
-    return ResultVo()
-
-
 if __name__ == "__main__":
     cors = config.get("cors", {})
     api.add_middleware(CORSMiddleware, allow_origins=cors.get("allow_origins", ["*"]),

client/client.py

@@ -15,9 +15,24 @@ from pydantic import AnyUrl
 from config.database import db_util
 from config.logger import logger
 from config.system import config
+from model.entity.AiChatRecommendQuestionEntity import AiChatRecommendQuestionEntity
 from model.entity.AiChatRecordEntity import AiChatRecordEntity
 
 
+async def record_answer(_id, _completion_answer):
+    """AI回答完成后，异步更新回答数据"""
+    await db_util.update(
+        AiChatRecordEntity,
+        {
+            "id": _id
+        },
+        {
+            "answer": _completion_answer,
+            "answer_time": datetime.now().strftime("%Y-%m-%d %H:%M:%S")
+        }
+    )
+
+
 class McpClient:
     def __init__(self):
         self.customer_resource: [] = None  # 客户资源
@@ -132,61 +147,79 @@ class McpClient:
         chat_record_entity = await db_util.add(chat_record_entity)
 
         logger.info(f"--> user origin query, message: {message}, customer_id: {customer_id}")
-        messages = [
-            {"role": "system", "content": self.default_system_prompt},
-            {"role": "system", "content": self.deal_customer_permission(customer_id)},
-            {"role": "system", "content": f"如果要使用到当前时间，请使用{datetime.now()}"},
-            {"role": "user", "content": message}
-        ]
-        logger.info(f"--> messages: {messages}")
-        logger.info(f"--> model: {self.model_name}")
-        # 调用ai
-        ai_response = await self.client.chat.completions.create(
-            model=self.model_name,
-            messages=messages,
-            tools=self.available_tools,
-            tool_choice="auto"
-        )
-        logger.info(f"--> ai response: {ai_response}")
-        chat_completion_message = ai_response.choices[0].message
-        logger.info(f"----> Available tools: {chat_completion_message.tool_calls}")
-        # 可能ai一次选取了多个工具，这里循环处理
-        for tool_call in chat_completion_message.tool_calls:
-            tool_name = tool_call.function.name
-            tool_args = json.loads(tool_call.function.arguments)
-            logger.info(f"------> start to call tool...")
-            logger.info(f"------> tool_name: {tool_name}, tool_args: {tool_args}")
-            result = await self.session.call_tool(tool_name, tool_args)
-            logger.info(f"------> call result: {result}")
-
-            messages.append(ChatCompletionAssistantMessageParam(
-                role="assistant",
-                tool_calls=[ChatCompletionMessageToolCallParam(
-                    id=tool_call.id,
-                    type="function",
-                    function=Function(
-                        name=tool_name,
-                        arguments=json.dumps(tool_args)
-                    )
-                )]
-            ))
-            messages.append(ChatCompletionToolMessageParam(
-                role="tool",
-                tool_call_id=tool_call.id,
-                content=str(result.content)
-            ))
-        ai_stream_response = await self.client.chat.completions.create(
-            model=self.model_name,
-            messages=messages,
-            tools=self.available_tools,
-            tool_choice="auto",
-            stream=True
+
+        db_question = await db_util.get_by_filter(
+            AiChatRecommendQuestionEntity,
+            AiChatRecommendQuestionEntity.question.__eq__(message),
         )
-        async for chunk in ai_stream_response:
-            if chunk.choices[0].finish_reason == "stop":
-                yield f"[DONE]-{chat_record_entity.id}"
-            else:
-                yield json.dumps({"content": chunk.choices[0].delta.content})
+        if db_question and len(db_question) > 0 and (default_answer := db_question[0].default_answer):
+            # 固定问题，模拟AI回答
+            step = 2  # 两个字符两个字符的输出
+            for i in range(0, len(default_answer), step):
+                await asyncio.sleep(0.05)  # 50毫秒延迟
+                yield json.dumps({"content": default_answer[i:i + step]})
+            yield "[DONE]"
+            asyncio.create_task(record_answer(chat_record_entity.id, default_answer))
+        else:
+            messages = [
+                {"role": "system", "content": self.default_system_prompt},
+                {"role": "system", "content": self.deal_customer_permission(customer_id)},
+                {"role": "system", "content": f"如果要使用到当前时间，请使用{datetime.now()}"},
+                {"role": "user", "content": message}
+            ]
+            logger.info(f"--> messages: {messages}")
+            logger.info(f"--> model: {self.model_name}")
+            # 调用ai
+            ai_response = await self.client.chat.completions.create(
+                model=self.model_name,
+                messages=messages,
+                tools=self.available_tools,
+                tool_choice="auto"
+            )
+            logger.info(f"--> ai response: {ai_response}")
+            chat_completion_message = ai_response.choices[0].message
+            logger.info(f"----> Available tools: {chat_completion_message.tool_calls}")
+            # 可能ai一次选取了多个工具，这里循环处理
+            for tool_call in chat_completion_message.tool_calls:
+                tool_name = tool_call.function.name
+                tool_args = json.loads(tool_call.function.arguments)
+                logger.info(f"------> start to call tool...")
+                logger.info(f"------> tool_name: {tool_name}, tool_args: {tool_args}")
+                result = await self.session.call_tool(tool_name, tool_args)
+                logger.info(f"------> call result: {result}")
+
+                messages.append(ChatCompletionAssistantMessageParam(
+                    role="assistant",
+                    tool_calls=[ChatCompletionMessageToolCallParam(
+                        id=tool_call.id,
+                        type="function",
+                        function=Function(
+                            name=tool_name,
+                            arguments=json.dumps(tool_args)
+                        )
+                    )]
+                ))
+                messages.append(ChatCompletionToolMessageParam(
+                    role="tool",
+                    tool_call_id=tool_call.id,
+                    content=str(result.content)
+                ))
+            ai_stream_response = await self.client.chat.completions.create(
+                model=self.model_name,
+                messages=messages,
+                tools=self.available_tools,
+                tool_choice="auto",
+                stream=True
+            )
+            completion_answer = ""
+            async for chunk in ai_stream_response:
+                if chunk.choices[0].finish_reason == "stop":
+                    asyncio.create_task(record_answer(chat_record_entity.id, completion_answer))
+                    yield "[DONE]"
+                else:
+                    content = chunk.choices[0].delta.content
+                    completion_answer += content
+                    yield json.dumps({"content": content})
 
 
 mcp_client_instance = McpClient()