2026-04-21 11:02:16 +08:00
|
|
|
|
"""
|
2026-05-05 17:30:55 +08:00
|
|
|
|
AI Agent 服务类 - 单图方案 + 动态模型选择
|
2026-04-21 11:02:16 +08:00
|
|
|
|
接收外部传入的 checkpointer,不负责管理连接生命周期
|
|
|
|
|
|
"""
|
|
|
|
|
|
|
|
|
|
|
|
import json
|
2026-04-26 17:44:52 +08:00
|
|
|
|
import asyncio
|
2026-05-05 17:30:55 +08:00
|
|
|
|
from typing import AsyncGenerator, Dict, Any, Optional, Tuple
|
2026-04-21 11:02:16 +08:00
|
|
|
|
|
2026-05-05 23:17:00 +08:00
|
|
|
|
# LangGraph 序列化器(修复 checkpoint 反序列化警告)
|
|
|
|
|
|
from langgraph.checkpoint.serde.jsonplus import JsonPlusSerializer
|
|
|
|
|
|
|
2026-04-21 11:02:16 +08:00
|
|
|
|
# 本地模块
|
2026-05-05 17:30:55 +08:00
|
|
|
|
from ..model_services import get_cached_chat_services
|
|
|
|
|
|
from ..main_graph.main_graph_builder import build_react_main_graph
|
2026-05-04 18:59:15 +08:00
|
|
|
|
from ..main_graph.tools.graph_tools import AVAILABLE_TOOLS, TOOLS_BY_NAME
|
|
|
|
|
|
from ..main_graph.config import set_stream_writer
|
|
|
|
|
|
from ..main_graph.utils.rag_initializer import init_rag_tool
|
2026-05-06 01:15:52 +08:00
|
|
|
|
from backend.app.core.intent_classifier import get_intent_classifier
|
|
|
|
|
|
from backend.app.logger import debug, info, warning, error
|
2026-05-04 18:59:15 +08:00
|
|
|
|
from ..main_graph.state import MainGraphState, CurrentAction
|
2026-04-21 11:02:16 +08:00
|
|
|
|
|
2026-05-01 11:24:13 +08:00
|
|
|
|
|
2026-05-05 23:17:00 +08:00
|
|
|
|
# ========== 自定义类型序列化器 ==========
|
|
|
|
|
|
def create_serde() -> JsonPlusSerializer:
|
|
|
|
|
|
"""创建带自定义类型注册的序列化器"""
|
|
|
|
|
|
from backend.app.core.intent import ReasoningAction, RetrievalConfig, ReasoningResult
|
|
|
|
|
|
|
|
|
|
|
|
return JsonPlusSerializer(
|
|
|
|
|
|
allowed_msgpack_modules=[
|
|
|
|
|
|
("app.core.intent", "ReasoningAction"),
|
|
|
|
|
|
("app.core.intent", "RetrievalConfig"),
|
|
|
|
|
|
("app.core.intent", "ReasoningResult"),
|
|
|
|
|
|
("app.main_graph.state", "CurrentAction"),
|
|
|
|
|
|
("app.main_graph.state", "ErrorSeverity"),
|
|
|
|
|
|
("app.main_graph.state", "ErrorRecord"),
|
|
|
|
|
|
]
|
|
|
|
|
|
)
|
|
|
|
|
|
|
|
|
|
|
|
|
2026-04-21 11:02:16 +08:00
|
|
|
|
class AIAgentService:
|
|
|
|
|
|
def __init__(self, checkpointer):
|
|
|
|
|
|
self.checkpointer = checkpointer
|
2026-05-05 17:30:55 +08:00
|
|
|
|
self.graph = None # 只有一张图
|
|
|
|
|
|
self.chat_services = None # 缓存的模型字典
|
2026-04-21 11:02:16 +08:00
|
|
|
|
self.tools = AVAILABLE_TOOLS.copy()
|
|
|
|
|
|
self.tools_by_name = TOOLS_BY_NAME.copy()
|
2026-04-26 17:37:57 +08:00
|
|
|
|
# 添加:意图分类器
|
|
|
|
|
|
self.intent_classifier = get_intent_classifier()
|
2026-04-26 17:44:52 +08:00
|
|
|
|
# RAG 管道(可选,需要时设置)
|
|
|
|
|
|
self.rag_pipeline = None
|
2026-05-01 15:43:45 +08:00
|
|
|
|
# Mem0 客户端
|
|
|
|
|
|
self.mem0_client = None
|
2026-04-21 11:02:16 +08:00
|
|
|
|
|
|
|
|
|
|
async def initialize(self):
|
2026-05-01 15:43:45 +08:00
|
|
|
|
# 0. 初始化 Mem0 客户端
|
2026-05-04 18:59:15 +08:00
|
|
|
|
from ..memory.mem0_client import Mem0Client
|
2026-05-05 13:30:31 +08:00
|
|
|
|
self.mem0_client = Mem0Client()
|
2026-05-01 15:43:45 +08:00
|
|
|
|
|
2026-04-21 11:02:16 +08:00
|
|
|
|
# 1. 初始化 RAG 工具(如果需要)
|
2026-05-05 13:30:31 +08:00
|
|
|
|
rag_tool = await init_rag_tool()
|
2026-04-21 11:02:16 +08:00
|
|
|
|
if rag_tool:
|
|
|
|
|
|
self.tools.append(rag_tool)
|
|
|
|
|
|
self.tools_by_name[rag_tool.name] = rag_tool
|
2026-05-05 04:32:42 +08:00
|
|
|
|
self.rag_tool = rag_tool # 保存到实例变量,供 config 注入
|
2026-05-05 17:30:55 +08:00
|
|
|
|
|
|
|
|
|
|
# 2. 获取缓存的模型字典
|
|
|
|
|
|
self.chat_services = get_cached_chat_services()
|
|
|
|
|
|
info(f"✅ 加载了 {len(self.chat_services)} 个可用模型: {list(self.chat_services.keys())}")
|
|
|
|
|
|
|
|
|
|
|
|
# 3. 只构建一次图(传入 chat_services 字典)
|
|
|
|
|
|
info(f"🔄 构建单图...")
|
|
|
|
|
|
graph_builder = build_react_main_graph(
|
|
|
|
|
|
chat_services=self.chat_services,
|
|
|
|
|
|
tools=self.tools,
|
|
|
|
|
|
mem0_client=self.mem0_client
|
|
|
|
|
|
)
|
2026-05-05 23:17:00 +08:00
|
|
|
|
# 注意:serde 已在创建 checkpointer 时传入,这里只需传入 checkpointer
|
2026-05-05 17:30:55 +08:00
|
|
|
|
self.graph = graph_builder.compile(checkpointer=self.checkpointer)
|
|
|
|
|
|
info(f"✅ 单图初始化完成")
|
|
|
|
|
|
|
2026-04-21 11:02:16 +08:00
|
|
|
|
return self
|
|
|
|
|
|
|
2026-05-05 17:30:55 +08:00
|
|
|
|
def _resolve_model(self, model: str) -> str:
|
|
|
|
|
|
"""
|
|
|
|
|
|
解析并验证模型名称,不可用时回退到第一个可用模型
|
|
|
|
|
|
|
|
|
|
|
|
Args:
|
|
|
|
|
|
model: 目标模型名称
|
|
|
|
|
|
|
|
|
|
|
|
Returns:
|
|
|
|
|
|
实际使用的模型名称
|
|
|
|
|
|
"""
|
|
|
|
|
|
if not model or model not in self.chat_services:
|
|
|
|
|
|
fallback = next(iter(self.chat_services.keys()))
|
|
|
|
|
|
warning(f"模型 '{model}' 不可用,回退到 '{fallback}'")
|
|
|
|
|
|
return fallback
|
|
|
|
|
|
return model
|
|
|
|
|
|
|
|
|
|
|
|
def _build_invocation(
|
|
|
|
|
|
self, message: str, thread_id: str, model: str, user_id: str
|
|
|
|
|
|
) -> Tuple[Dict[str, Any], Dict[str, Any]]:
|
|
|
|
|
|
"""
|
|
|
|
|
|
构建图调用所需的 config 和 input_state
|
|
|
|
|
|
|
|
|
|
|
|
Args:
|
|
|
|
|
|
message: 用户消息
|
|
|
|
|
|
thread_id: 会话 ID
|
|
|
|
|
|
model: 模型名称
|
|
|
|
|
|
user_id: 用户 ID
|
|
|
|
|
|
|
|
|
|
|
|
Returns:
|
|
|
|
|
|
(config, input_state) 元组
|
|
|
|
|
|
"""
|
2026-04-21 11:02:16 +08:00
|
|
|
|
config = {
|
2026-05-05 04:32:42 +08:00
|
|
|
|
"configurable": {
|
|
|
|
|
|
"thread_id": thread_id,
|
2026-05-05 17:30:55 +08:00
|
|
|
|
"rag_tool": getattr(self, "rag_tool", None),
|
2026-05-05 04:32:42 +08:00
|
|
|
|
},
|
2026-04-21 11:02:16 +08:00
|
|
|
|
"metadata": {"user_id": user_id}
|
|
|
|
|
|
}
|
2026-05-01 00:13:13 +08:00
|
|
|
|
input_state = {
|
|
|
|
|
|
"user_query": message,
|
|
|
|
|
|
"messages": [{"role": "user", "content": message}],
|
|
|
|
|
|
"user_id": user_id,
|
2026-05-05 17:30:55 +08:00
|
|
|
|
"current_model": model,
|
2026-05-01 00:13:13 +08:00
|
|
|
|
"current_action": CurrentAction.NONE
|
|
|
|
|
|
}
|
2026-05-05 17:30:55 +08:00
|
|
|
|
return config, input_state
|
2026-04-21 11:02:16 +08:00
|
|
|
|
|
2026-05-05 17:30:55 +08:00
|
|
|
|
async def process_message(
|
|
|
|
|
|
self, message: str, thread_id: str, model: str = "", user_id: str = "default_user"
|
|
|
|
|
|
) -> dict:
|
|
|
|
|
|
"""处理用户消息,返回包含回复、token统计和耗时的字典"""
|
|
|
|
|
|
# 解析模型名称
|
|
|
|
|
|
resolved_model = self._resolve_model(model)
|
|
|
|
|
|
|
|
|
|
|
|
# 构建调用参数
|
|
|
|
|
|
config, input_state = self._build_invocation(message, thread_id, resolved_model, user_id)
|
|
|
|
|
|
|
|
|
|
|
|
result = await self.graph.ainvoke(input_state, config=config)
|
2026-04-21 11:02:16 +08:00
|
|
|
|
|
2026-05-01 00:13:13 +08:00
|
|
|
|
reply = result.get("final_result", "")
|
|
|
|
|
|
if not reply and result.get("messages"):
|
|
|
|
|
|
reply = result["messages"][-1].content
|
2026-05-05 17:30:55 +08:00
|
|
|
|
token_usage = result.get("last_token_usage", {})
|
|
|
|
|
|
elapsed_time = result.get("last_elapsed_time", 0.0)
|
|
|
|
|
|
actual_model = result.get("current_model", resolved_model)
|
2026-04-21 11:02:16 +08:00
|
|
|
|
|
|
|
|
|
|
return {
|
|
|
|
|
|
"reply": reply,
|
|
|
|
|
|
"token_usage": token_usage,
|
2026-05-05 17:30:55 +08:00
|
|
|
|
"elapsed_time": elapsed_time,
|
|
|
|
|
|
"model_used": actual_model
|
2026-04-21 11:02:16 +08:00
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
def _serialize_value(self, value):
|
|
|
|
|
|
"""递归将 LangChain 对象转换为可 JSON 序列化的格式"""
|
|
|
|
|
|
if hasattr(value, 'content'):
|
|
|
|
|
|
msg_type = getattr(value, 'type', 'message')
|
|
|
|
|
|
return {
|
|
|
|
|
|
"role": msg_type,
|
|
|
|
|
|
"content": getattr(value, 'content', ''),
|
|
|
|
|
|
"additional_kwargs": getattr(value, 'additional_kwargs', {}),
|
|
|
|
|
|
"tool_calls": getattr(value, 'tool_calls', [])
|
|
|
|
|
|
}
|
|
|
|
|
|
elif isinstance(value, dict):
|
|
|
|
|
|
return {k: self._serialize_value(v) for k, v in value.items()}
|
|
|
|
|
|
elif isinstance(value, (list, tuple)):
|
|
|
|
|
|
return [self._serialize_value(item) for item in value]
|
|
|
|
|
|
else:
|
|
|
|
|
|
try:
|
|
|
|
|
|
json.dumps(value)
|
|
|
|
|
|
return value
|
|
|
|
|
|
except (TypeError, ValueError):
|
|
|
|
|
|
return str(value)
|
|
|
|
|
|
|
2026-05-05 17:30:55 +08:00
|
|
|
|
async def _handle_message_chunk(
|
|
|
|
|
|
self, chunk: Dict[str, Any], current_node: Optional[str], tool_calls_in_progress: Dict[str, Any]
|
|
|
|
|
|
) -> AsyncGenerator[Dict[str, Any], None]:
|
|
|
|
|
|
"""处理 messages 类型的 chunk"""
|
|
|
|
|
|
message_chunk, metadata = chunk["data"]
|
|
|
|
|
|
node_name = metadata.get("langgraph_node", "unknown")
|
|
|
|
|
|
new_current_node = current_node
|
2026-04-21 11:02:16 +08:00
|
|
|
|
|
2026-05-05 17:30:55 +08:00
|
|
|
|
# 检测节点变化,发送节点开始事件
|
|
|
|
|
|
if node_name != current_node:
|
|
|
|
|
|
if current_node:
|
|
|
|
|
|
yield {"type": "node_end", "node": current_node}
|
|
|
|
|
|
yield {"type": "node_start", "node": node_name}
|
|
|
|
|
|
new_current_node = node_name
|
|
|
|
|
|
|
|
|
|
|
|
# 处理消息内容
|
|
|
|
|
|
token_content = getattr(message_chunk, 'content', str(message_chunk))
|
|
|
|
|
|
reasoning_token = ""
|
|
|
|
|
|
if hasattr(message_chunk, 'additional_kwargs'):
|
|
|
|
|
|
reasoning_token = message_chunk.additional_kwargs.get("reasoning_content", "")
|
|
|
|
|
|
|
|
|
|
|
|
# 处理思考过程
|
|
|
|
|
|
if reasoning_token:
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "llm_token",
|
|
|
|
|
|
"node": node_name,
|
|
|
|
|
|
"reasoning_token": reasoning_token
|
|
|
|
|
|
}
|
|
|
|
|
|
# 处理工具调用
|
|
|
|
|
|
elif hasattr(message_chunk, 'tool_calls') and message_chunk.tool_calls:
|
|
|
|
|
|
for tool_call in message_chunk.tool_calls:
|
|
|
|
|
|
tool_call_id = tool_call.get("id", "")
|
|
|
|
|
|
tool_name = tool_call.get("name", "")
|
|
|
|
|
|
tool_args = tool_call.get("args", {})
|
|
|
|
|
|
|
|
|
|
|
|
# 记录工具调用开始,避免重复
|
|
|
|
|
|
if tool_call_id and tool_call_id not in tool_calls_in_progress:
|
|
|
|
|
|
tool_calls_in_progress[tool_call_id] = {
|
|
|
|
|
|
"name": tool_name,
|
|
|
|
|
|
"args": tool_args
|
|
|
|
|
|
}
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "tool_call_start",
|
|
|
|
|
|
"tool": tool_name,
|
|
|
|
|
|
"args": tool_args,
|
|
|
|
|
|
"id": tool_call_id
|
|
|
|
|
|
}
|
|
|
|
|
|
# 处理普通 token
|
|
|
|
|
|
elif token_content:
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "llm_token",
|
|
|
|
|
|
"node": node_name,
|
|
|
|
|
|
"token": token_content,
|
|
|
|
|
|
"reasoning_token": reasoning_token
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
# 返回更新后的 current_node
|
|
|
|
|
|
yield {"type": "_update_state", "current_node": new_current_node}
|
|
|
|
|
|
|
|
|
|
|
|
async def _handle_updates_chunk(
|
|
|
|
|
|
self, chunk: Dict[str, Any], tool_calls_in_progress: Dict[str, Any], actual_model_used: str
|
|
|
|
|
|
) -> AsyncGenerator[Dict[str, Any], None]:
|
|
|
|
|
|
"""处理 updates 类型的 chunk"""
|
|
|
|
|
|
updates_data = chunk["data"]
|
|
|
|
|
|
new_actual_model = actual_model_used
|
|
|
|
|
|
|
|
|
|
|
|
debug(f"[Stream] updates 数据: {list(updates_data.keys()) if isinstance(updates_data, dict) else type(updates_data)}")
|
|
|
|
|
|
|
|
|
|
|
|
# 特别检查 final_result 和 current_model
|
|
|
|
|
|
if isinstance(updates_data, dict):
|
|
|
|
|
|
if "final_result" in updates_data:
|
|
|
|
|
|
debug(f"[Stream] 收到 final_result: {str(updates_data['final_result'])[:100]}...")
|
|
|
|
|
|
if "current_model" in updates_data:
|
|
|
|
|
|
new_actual_model = updates_data["current_model"]
|
|
|
|
|
|
info(f"[Stream] 实际使用模型: {new_actual_model}")
|
|
|
|
|
|
|
|
|
|
|
|
serialized_data = self._serialize_value(updates_data)
|
|
|
|
|
|
|
|
|
|
|
|
# 检查是否有人工审核请求
|
|
|
|
|
|
if "review_pending" in serialized_data and serialized_data["review_pending"]:
|
|
|
|
|
|
review_id = serialized_data.get("review_id", "")
|
|
|
|
|
|
content_to_review = serialized_data.get("content_to_review", "")
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "human_review_request",
|
|
|
|
|
|
"review_id": review_id,
|
|
|
|
|
|
"content": content_to_review
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
# 检查是否有工具结果
|
|
|
|
|
|
if "messages" in serialized_data:
|
|
|
|
|
|
for msg in serialized_data["messages"]:
|
|
|
|
|
|
# 检测工具结果消息
|
|
|
|
|
|
if msg.get("role") == "tool":
|
|
|
|
|
|
tool_call_id = msg.get("tool_call_id", "")
|
|
|
|
|
|
tool_name = msg.get("name", "")
|
|
|
|
|
|
tool_result = msg.get("content", "")
|
|
|
|
|
|
|
|
|
|
|
|
if tool_call_id and tool_call_id in tool_calls_in_progress:
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "tool_call_end",
|
|
|
|
|
|
"tool": tool_name,
|
|
|
|
|
|
"id": tool_call_id,
|
|
|
|
|
|
"result": tool_result
|
|
|
|
|
|
}
|
|
|
|
|
|
del tool_calls_in_progress[tool_call_id]
|
|
|
|
|
|
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "state_update",
|
|
|
|
|
|
"data": serialized_data
|
2026-05-01 00:54:58 +08:00
|
|
|
|
}
|
2026-05-05 17:30:55 +08:00
|
|
|
|
|
|
|
|
|
|
# 返回更新后的模型
|
|
|
|
|
|
yield {"type": "_update_state", "actual_model_used": new_actual_model}
|
2026-04-21 11:02:16 +08:00
|
|
|
|
|
2026-05-05 17:30:55 +08:00
|
|
|
|
async def _handle_custom_chunk(self, chunk: Dict[str, Any]) -> AsyncGenerator[Dict[str, Any], None]:
|
|
|
|
|
|
"""处理 custom 类型的 chunk"""
|
|
|
|
|
|
custom_data = chunk["data"]
|
|
|
|
|
|
|
|
|
|
|
|
# 处理我们从 react_reason_node 发送的自定义推理事件
|
|
|
|
|
|
if isinstance(custom_data, dict):
|
|
|
|
|
|
# 检查是否是我们的推理事件
|
|
|
|
|
|
if "action" in custom_data and "reasoning" in custom_data:
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "react_reasoning",
|
|
|
|
|
|
"step": custom_data.get("step", 1),
|
|
|
|
|
|
"action": custom_data.get("action", "unknown"),
|
|
|
|
|
|
"confidence": custom_data.get("confidence", 0),
|
|
|
|
|
|
"reasoning": custom_data.get("reasoning", "")
|
|
|
|
|
|
}
|
|
|
|
|
|
else:
|
|
|
|
|
|
# 处理其他自定义事件
|
|
|
|
|
|
serialized_data = self._serialize_value(custom_data)
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "custom",
|
|
|
|
|
|
"data": serialized_data
|
|
|
|
|
|
}
|
|
|
|
|
|
else:
|
|
|
|
|
|
# 处理其他自定义事件
|
|
|
|
|
|
serialized_data = self._serialize_value(custom_data)
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "custom",
|
|
|
|
|
|
"data": serialized_data
|
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
|
|
async def process_message_stream(
|
|
|
|
|
|
self, message: str, thread_id: str, model: str = "", user_id: str = "default_user"
|
|
|
|
|
|
) -> AsyncGenerator[Dict[str, Any], None]:
|
|
|
|
|
|
"""流式处理消息,返回异步生成器"""
|
|
|
|
|
|
# 解析模型名称
|
|
|
|
|
|
resolved_model = self._resolve_model(model)
|
|
|
|
|
|
|
|
|
|
|
|
# 构建调用参数
|
|
|
|
|
|
config, input_state = self._build_invocation(message, thread_id, resolved_model, user_id)
|
|
|
|
|
|
|
|
|
|
|
|
# ========== 意图识别(保留用于日志和后续路由)==========
|
2026-04-26 17:44:52 +08:00
|
|
|
|
intent_result = await self.intent_classifier.classify(message)
|
|
|
|
|
|
info(f"🧠 意图识别: {intent_result.intent_type} (置信度: {intent_result.confidence:.2f})")
|
|
|
|
|
|
info(f"📝 推理: {intent_result.reasoning}")
|
|
|
|
|
|
|
2026-05-05 17:30:55 +08:00
|
|
|
|
# 注入意图到状态(让 hybrid_router 可以利用)
|
|
|
|
|
|
input_state["intent_type"] = intent_result.intent_type.value
|
|
|
|
|
|
input_state["intent_confidence"] = intent_result.confidence
|
|
|
|
|
|
|
2026-04-26 17:44:52 +08:00
|
|
|
|
# 发送意图分类事件
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "intent_classified",
|
|
|
|
|
|
"intent": intent_result.intent_type.value,
|
|
|
|
|
|
"confidence": intent_result.confidence,
|
|
|
|
|
|
"reasoning": intent_result.reasoning
|
|
|
|
|
|
}
|
|
|
|
|
|
|
2026-05-05 17:30:55 +08:00
|
|
|
|
# 发送路径决策事件(目前硬编码,但状态中有意图信息供后续使用)
|
2026-04-26 17:44:52 +08:00
|
|
|
|
yield {
|
|
|
|
|
|
"type": "path_decision",
|
2026-05-01 11:24:13 +08:00
|
|
|
|
"path": "react_loop",
|
2026-04-26 17:44:52 +08:00
|
|
|
|
"intent": intent_result.intent_type.value
|
|
|
|
|
|
}
|
2026-05-05 17:30:55 +08:00
|
|
|
|
# =============================================
|
2026-05-01 11:24:13 +08:00
|
|
|
|
|
|
|
|
|
|
# ========== React 循环路径 ==========
|
2026-05-05 17:30:55 +08:00
|
|
|
|
info(f"🚀 开始执行单图,指定模型: {resolved_model}")
|
2026-05-01 11:24:13 +08:00
|
|
|
|
current_node = None
|
2026-05-05 17:30:55 +08:00
|
|
|
|
tool_calls_in_progress: Dict[str, Any] = {}
|
|
|
|
|
|
actual_model_used = resolved_model
|
|
|
|
|
|
chunk_count = 0
|
|
|
|
|
|
full_message_content = ""
|
2026-05-01 11:24:13 +08:00
|
|
|
|
|
2026-05-01 12:11:16 +08:00
|
|
|
|
try:
|
|
|
|
|
|
info(f"📡 开始调用 graph.astream()...")
|
2026-05-04 04:28:32 +08:00
|
|
|
|
|
2026-05-05 17:30:55 +08:00
|
|
|
|
async for chunk in self.graph.astream(
|
2026-05-01 12:11:16 +08:00
|
|
|
|
input_state,
|
|
|
|
|
|
config=config,
|
|
|
|
|
|
stream_mode=["messages", "updates", "custom"],
|
|
|
|
|
|
version="v2",
|
|
|
|
|
|
subgraphs=True
|
|
|
|
|
|
):
|
|
|
|
|
|
chunk_count += 1
|
|
|
|
|
|
chunk_type = chunk["type"]
|
|
|
|
|
|
|
|
|
|
|
|
if chunk_type == "messages":
|
2026-05-05 17:30:55 +08:00
|
|
|
|
async for event in self._handle_message_chunk(
|
|
|
|
|
|
chunk, current_node, tool_calls_in_progress
|
|
|
|
|
|
):
|
|
|
|
|
|
if event.get("type") == "_update_state":
|
|
|
|
|
|
current_node = event.get("current_node", current_node)
|
|
|
|
|
|
else:
|
|
|
|
|
|
# 如果是 llm_call 节点的 token,收集完整消息
|
|
|
|
|
|
if (
|
|
|
|
|
|
event.get("type") == "llm_token"
|
|
|
|
|
|
and event.get("node") == "llm_call"
|
|
|
|
|
|
and "token" in event
|
|
|
|
|
|
):
|
|
|
|
|
|
full_message_content += event["token"]
|
|
|
|
|
|
yield event
|
2026-04-26 17:44:52 +08:00
|
|
|
|
|
2026-05-01 12:11:16 +08:00
|
|
|
|
elif chunk_type == "updates":
|
2026-05-05 17:30:55 +08:00
|
|
|
|
async for event in self._handle_updates_chunk(
|
|
|
|
|
|
chunk, tool_calls_in_progress, actual_model_used
|
|
|
|
|
|
):
|
|
|
|
|
|
if event.get("type") == "_update_state":
|
|
|
|
|
|
actual_model_used = event.get("actual_model_used", actual_model_used)
|
2026-05-02 09:39:18 +08:00
|
|
|
|
else:
|
2026-05-05 17:30:55 +08:00
|
|
|
|
yield event
|
2026-04-26 16:05:44 +08:00
|
|
|
|
|
2026-05-05 17:30:55 +08:00
|
|
|
|
elif chunk_type == "custom":
|
|
|
|
|
|
async for event in self._handle_custom_chunk(chunk):
|
|
|
|
|
|
yield event
|
2026-04-26 16:05:44 +08:00
|
|
|
|
|
2026-05-04 04:28:32 +08:00
|
|
|
|
# 完整消息集合完成后,一次性打印
|
2026-05-05 17:30:55 +08:00
|
|
|
|
info(f"✅ graph.astream() 完成,共 {chunk_count} 个 chunks")
|
2026-05-04 04:28:32 +08:00
|
|
|
|
if full_message_content:
|
|
|
|
|
|
info(f"📄 完整消息内容: {repr(full_message_content)}")
|
2026-05-05 17:30:55 +08:00
|
|
|
|
info(f"🤖 实际使用模型: {actual_model_used}")
|
2026-04-26 16:05:44 +08:00
|
|
|
|
|
2026-05-01 12:11:16 +08:00
|
|
|
|
except Exception as e:
|
2026-05-05 17:30:55 +08:00
|
|
|
|
error(f"❌ 执行单图时出错: {e}")
|
2026-05-01 12:11:16 +08:00
|
|
|
|
import traceback
|
|
|
|
|
|
error(f"📋 堆栈: {traceback.format_exc()}")
|
2026-04-26 16:05:44 +08:00
|
|
|
|
yield {
|
2026-05-05 17:30:55 +08:00
|
|
|
|
"type": "error",
|
|
|
|
|
|
"message": str(e)
|
|
|
|
|
|
}
|
|
|
|
|
|
finally:
|
|
|
|
|
|
# 无论成功或失败,都发送结束事件,保证前端平稳关闭
|
|
|
|
|
|
if current_node:
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "node_end",
|
|
|
|
|
|
"node": current_node
|
|
|
|
|
|
}
|
|
|
|
|
|
yield {
|
|
|
|
|
|
"type": "done",
|
|
|
|
|
|
"model_used": actual_model_used
|
2026-04-26 16:05:44 +08:00
|
|
|
|
}
|