PicoBot/src/gateway/processor.rs
oudecheng b042b45ac7 fix: 话题隔离 - topic_id 全程显式传递,修复消息错投与并发阻塞
问题:WS 端一个话题执行中时新建另一话题发消息,新会话无响应
(串行锁按 chat_id 阻塞),且首个话题完成后用户消息错误出现在
旧话题而非新话题(执行路径多次从共享 UI 状态读取 topic_id 产生竞态)。

核心修复(第一性原则):
执行上下文应在消息接收时一次性捕获,全程显式传递,不从共享可变
状态重复读取。

1. processor.rs: process_one 入口捕获 current_topic,传入 handle_message
   和 set_agent_cancel_token
2. session_message_service.rs: handle_message 签名加 topic_id 参数,
   透传给 MessageExecutionRequest
3. execution.rs: MessageExecutionRequest 加 topic_id 字段;
   - 串行锁键改用 topic_id(不同 topic 并发,同 topic 串行)
   - original_topic_id 优先用传入值,消除锁等待期间 topic 切换竞态
   - append_persisted_message 调用传入 original_topic_id
   - create_agent 调用传入 original_topic_id
4. session.rs: append_persisted_message 加 explicit_topic_id 参数;
   create_agent/create_agent_with_provider_config 加 explicit_topic_id;
   set_cancel_receiver/set_agent_cancel_token 加 topic_id 参数;
   pending_cancel_tokens 查找改为优先 topic_id(避免并发 topic 执行时
   cancel token 互相覆盖)

对抗性审查补丁:
- append_persisted_message: 仅当写入 topic 匹配当前活跃 topic 时才更新
  内存历史,避免旧 topic 的排队消息污染已切换到的新 topic 内存历史
- prepare_and_execute_scheduled_task: 锁前一次性捕获 topic_id,锁后
  复用同一值作为 original_topic_id,保证锁键与写入目标一致

已验证:cargo check 通过,gateway 模块 48 个测试通过(1 个预存在的
prompt 模板测试失败,与本次修改无关)。
2026-07-28 14:19:20 +08:00

411 lines
18 KiB
Rust
Raw Blame History

This file contains ambiguous Unicode characters

This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

use std::collections::HashSet;
use std::sync::{Arc, Mutex};
use tokio::sync::Semaphore;
use crate::agent::{AgentError, PersistingEmittedMessageHandler};
use crate::bus::{InboundMessage, MessageBus, OutboundMessage};
use crate::command::adapter::InputAdapter;
use crate::command::adapters::channel::ChannelInputAdapter;
use crate::command::handler::CommandRouter;
use crate::command::handlers::delete_topic::DeleteTopicCommandHandler;
use crate::command::handlers::get_current::GetCurrentSessionCommandHandler;
use crate::command::handlers::help::HelpCommandHandler;
use crate::command::handlers::list_sessions::ListSessionsCommandHandler;
use crate::command::handlers::load_topic::LoadTopicCommandHandler;
use crate::command::handlers::save_session::SaveSessionCommandHandler;
use crate::command::handlers::save_topic::SaveTopicCommandHandler;
use crate::command::handlers::session::SessionCommandHandler;
use crate::command::handlers::stop_execution::StopExecutionCommandHandler;
use crate::command::handlers::switch_topic::SwitchTopicCommandHandler;
use crate::config::LLMProviderConfig;
use crate::gateway::agent_factory::build_system_prompt_provider;
use crate::gateway::cancel_manager::CancelManager;
use crate::providers::{create_provider, ProviderRuntimeConfig};
use crate::storage::persistent_session_id;
use crate::topic_description::generate_topic_description;
use super::session::{BusToolCallEmitter, SessionManager};
#[derive(Clone)]
pub struct InboundProcessor {
bus: Arc<MessageBus>,
session_manager: SessionManager,
semaphore: Arc<Semaphore>,
provider_config: LLMProviderConfig,
command_router: Arc<CommandRouter>,
cancel_manager: CancelManager,
description_generation_in_flight: Arc<Mutex<HashSet<String>>>,
}
impl InboundProcessor {
pub fn new(
bus: Arc<MessageBus>,
session_manager: SessionManager,
semaphore: Arc<Semaphore>,
provider_config: LLMProviderConfig,
cancel_manager: CancelManager,
) -> Self {
// 创建命令路由器并注册处理器
let mut command_router = CommandRouter::new();
let store = session_manager.store();
// 注册 Session 处理器
let session_handler = SessionCommandHandler::new(store.clone())
.with_session_manager(session_manager.clone());
command_router.register(Box::new(session_handler));
// 注册 list_sessions 处理器
command_router.register(Box::new(ListSessionsCommandHandler::new(store.clone())));
// 注册 switch_topic 处理器
let switch_handler = SwitchTopicCommandHandler::new(store.clone())
.with_session_manager(session_manager.clone());
command_router.register(Box::new(switch_handler));
// 创建 system_prompt_provider用于 save_session, save_topic, get_current
// 与 AgentFactory::create 共享同一构建逻辑,确保保存到文件的系统提示词
// 与 LLM 实际接收的提示词完全一致(含 Expert/Subagent/Todo
let system_prompt_provider = build_system_prompt_provider(
0, // 命令侧不需要 reinject 逻辑
provider_config.clone(),
session_manager.store().clone(),
session_manager.skills(),
session_manager.experts(),
session_manager.subagent_runtime(),
);
// 注册 get_current 处理器
command_router.register(Box::new(
GetCurrentSessionCommandHandler::new(store.clone())
.with_system_prompt_provider(system_prompt_provider.clone())
));
// 注册 load_topic 处理器
command_router.register(Box::new(LoadTopicCommandHandler::new(store.clone())));
// 注册 save_session 处理器
command_router.register(Box::new(SaveSessionCommandHandler::new(
store.clone(),
session_manager.task_repository(),
system_prompt_provider.clone(),
)));
// 注册 save_topic 处理器
command_router.register(Box::new(SaveTopicCommandHandler::new(
store.clone(),
session_manager.task_repository(),
system_prompt_provider,
)));
// 注册 delete_topic 处理器
command_router.register(Box::new(
DeleteTopicCommandHandler::new(store.clone())
.with_session_manager(session_manager.clone()),
));
// 注册 help 处理器(最后注册,获取所有已注册命令的元数据)
let metadata = command_router.metadata_arc();
command_router.register(Box::new(HelpCommandHandler::new(metadata)));
// 注册 stop_execution 处理器
command_router.register(Box::new(StopExecutionCommandHandler::new(
cancel_manager.clone(),
session_manager.clone(),
)));
Self {
bus,
session_manager,
semaphore,
provider_config,
command_router: Arc::new(command_router),
cancel_manager,
description_generation_in_flight: Arc::new(Mutex::new(HashSet::new())),
}
}
pub async fn run(self) {
let max_concurrent = self.semaphore.available_permits();
tracing::info!(
max_concurrent_requests = max_concurrent,
"Inbound processor started"
);
loop {
// 1. 消费消息 (channel 关闭时返回 None优雅退出)
let inbound = match self.bus.consume_inbound().await {
Some(msg) => msg,
None => {
tracing::info!("Inbound bus closed, stopping inbound processor");
break;
}
};
#[cfg(debug_assertions)]
{
tracing::debug!(
channel = %inbound.channel,
chat_id = %inbound.chat_id,
sender = %inbound.sender_id,
content_len = %inbound.content.len(),
media_count = %inbound.media.len(),
"Processing inbound message"
);
}
// 2. 获取 semaphore permit控制并发
let permit = match self.semaphore.clone().acquire_owned().await {
Ok(permit) => permit,
Err(_) => {
tracing::error!("Semaphore closed, stopping inbound processor");
break;
}
};
// 3. 克隆 processor 用于新任务
let processor = self.clone();
// 4. 独立任务处理(包含 permit任务完成自动释放
tokio::spawn(async move {
let _permit = permit; // 持有 permit 直到任务完成
if let Err(e) = processor.process_one(inbound).await {
tracing::error!(error = %e, "Message processing failed");
}
});
}
}
async fn process_one(&self, inbound: InboundMessage) -> Result<(), AgentError> {
// 计算正确的 session_id根据 channel_name 和 chat_id
let session_id = persistent_session_id(&inbound.channel, &inbound.chat_id);
// 获取当前话题(封装了 session 创建逻辑)
let current_topic = self.session_manager
.get_current_topic(&inbound.channel, &inbound.chat_id)
.await?;
// 使用 ChannelInputAdapter 尝试解析命令
let adapter = ChannelInputAdapter::new();
let ctx = crate::command::context::AdapterContext::new(&inbound.channel)
.with_session_id(&session_id);
if let Ok(Some(cmd)) = adapter.try_parse(&inbound.content, ctx) {
// 使用命令路由器处理
let mut cmd_ctx = crate::command::context::CommandContext::new(&inbound.channel, &inbound.channel)
.with_session_id(&session_id)
.with_chat_id(&inbound.chat_id);
// 只在有话题时才设置 topic_id
if let Some(ref topic_id) = current_topic {
cmd_ctx = cmd_ctx.with_topic_id(topic_id.as_str());
}
let response = self.command_router.dispatch_with_response(cmd, cmd_ctx).await;
// 发送响应给用户
if response.success {
// 提取响应消息
// chat_id 保持为 inbound.chat_id飞书 open_id
// session_id 放入 metadata 用于会话管理
for msg in &response.messages {
if let Err(error) = self
.bus
.publish_outbound(OutboundMessage::assistant(
inbound.channel.clone(),
inbound.chat_id.clone(),
response.metadata.get("session_id").cloned(),
msg.content.clone(),
None,
inbound.forwarded_metadata.clone(),
))
.await
{
tracing::error!(error = %error, "Failed to publish command response");
}
}
} else if let Some(error) = response.error {
if let Err(e) = self
.bus
.publish_outbound(OutboundMessage::assistant(
inbound.channel.clone(),
inbound.chat_id.clone(),
response.metadata.get("session_id").cloned(),
format!("Error [{}]: {}", error.code, error.message),
None,
inbound.forwarded_metadata.clone(),
))
.await
{
tracing::error!(error = %e, "Failed to publish error response");
}
}
return Ok(());
}
// 普通消息进入 AgentLoop
// 构建 emitter metadata包含 forwarded_metadata 和 topic_id用于前端消息隔离
let mut emitter_metadata = inbound.forwarded_metadata.clone();
if let Some(ref topic_id) = current_topic {
emitter_metadata.insert("topic_id".to_string(), topic_id.clone());
}
let live_emitter = Arc::new(PersistingEmittedMessageHandler::new(
BusToolCallEmitter::new(
self.bus.clone(),
inbound.channel.clone(),
inbound.chat_id.clone(),
emitter_metadata,
self.session_manager.store(),
),
self.session_manager.store(),
&session_id,
current_topic.clone(),
));
// 保存 channel 和 chat_id 用于后续清理(因 match 中可能 move inbound
let channel = inbound.channel.clone();
let chat_id = inbound.chat_id.clone();
// 按 topic_id 注册取消信号Agent 构建时通过 Session 消费该 receiver
if let Some(ref topic_id) = current_topic {
let cancel_rx = self.cancel_manager.register(topic_id).await;
self.session_manager
.set_agent_cancel_token(&channel, &chat_id, Some(topic_id.as_str()), cancel_rx)
.await;
}
match self
.session_manager
.handle_message(
&inbound.channel,
&inbound.sender_id,
&inbound.chat_id,
&inbound.content,
inbound.media,
Some(live_emitter),
current_topic.as_deref(),
)
.await
{
Ok(outbound_messages) => {
for mut outbound in outbound_messages {
outbound.metadata.extend(inbound.forwarded_metadata.clone());
// 注入 topic_id 到 outbound metadata用于前端按话题隔离消息
if let Some(ref topic_id) = current_topic {
outbound.metadata.insert("topic_id".to_string(), topic_id.clone());
}
if let Err(error) = self.bus.publish_outbound(outbound).await {
tracing::error!(error = %error, "Failed to publish outbound");
}
}
// 异步生成 topic 描述(仅当描述为空且没有正在进行的生成任务时触发)
if let Some(ref topic_id) = current_topic {
let store = self.session_manager.store();
if let Ok(Some(topic)) = store.get_topic(topic_id) {
if topic.description.is_none() || topic.description.as_ref().map(|d| d.is_empty()).unwrap_or(true) {
// 检查并设置"生成中"守卫,防止竞态条件导致重复生成
let should_generate = {
let mut in_flight = self.description_generation_in_flight.lock().unwrap();
if in_flight.contains(topic_id) {
false
} else {
in_flight.insert(topic_id.clone());
true
}
};
if should_generate {
let provider_config = self.provider_config.clone();
let topic_id_clone = topic_id.clone();
let store_clone = store.clone();
let in_flight = self.description_generation_in_flight.clone();
tokio::spawn(async move {
// 从 DB 查询该 topic 的第一条用户消息作为描述生成的依据
let first_user_message = store_clone
.load_messages_for_topic(&topic_id_clone, None)
.ok()
.and_then(|msgs| msgs.into_iter().find(|m| m.role == "user"))
.map(|m| m.content);
let message_content = match first_user_message {
Some(content) => content,
None => {
tracing::warn!(topic_id = %topic_id_clone, "No user message found for topic, skipping description generation");
in_flight.lock().unwrap().remove(&topic_id_clone);
return;
}
};
let runtime_config: ProviderRuntimeConfig = provider_config.into();
if let Ok(provider) = create_provider(runtime_config) {
match generate_topic_description(provider.as_ref(), &message_content).await {
Ok(description) => {
if let Err(e) = store_clone.update_topic_description(&topic_id_clone, &description) {
tracing::error!(error = %e, topic_id = %topic_id_clone, "Failed to update topic description");
} else {
tracing::info!(topic_id = %topic_id_clone, description = %description, "Topic description generated");
}
}
Err(e) => {
tracing::error!(error = %e, topic_id = %topic_id_clone, "Failed to generate topic description");
}
}
}
// 无论成功失败,释放生成守卫
in_flight.lock().unwrap().remove(&topic_id_clone);
});
}
}
}
}
}
Err(error) => {
tracing::error!(error = %error, "Failed to handle message");
let mut metadata = inbound.forwarded_metadata.clone();
metadata.insert("error_kind".to_string(), "agent_execution".to_string());
if let Err(publish_error) = self
.bus
.publish_outbound(OutboundMessage::error_notification(
inbound.channel,
inbound.chat_id,
None, // session_id
error.to_string(),
None,
metadata,
))
.await
{
tracing::error!(error = %publish_error, "Failed to publish execution error outbound");
}
}
}
// 清理取消信号注册(幂等:如果已被 cancel_by_topic() 移除则为 no-op
if let Some(ref topic_id) = current_topic {
self.cancel_manager.remove_by_topic(topic_id).await;
}
// 发送执行完成信号,通知前端可以停止 loading 状态
// 无论成功还是失败都发送,确保前端状态正确
let mut completion_metadata = inbound.forwarded_metadata.clone();
if let Some(ref topic_id) = current_topic {
completion_metadata.insert("topic_id".to_string(), topic_id.clone());
}
if let Err(error) = self
.bus
.publish_outbound(OutboundMessage::execution_completed(
channel,
chat_id,
Some(session_id),
completion_metadata,
))
.await
{
tracing::error!(error = %error, "Failed to publish execution_completed");
}
Ok(())
}
}