fix: propagate real message IDs via SSE and fix regenerate model parameter

- Add message_id and user_message_id to SSE done events so frontend
  can replace temp IDs with real DB IDs, fixing "消息不存在" on regenerate
- Replace Body(default=None) with RegenerateRequest Pydantic model for
  proper JSON body parsing, fixing 422 Unprocessable Content
- Frontend always sends model in regenerate request body
- Pass selected model through message-item → regenerateMessage chain
- Various UI refinements to chat sidebar, quick questions, and layout

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>

web/src/store/chat.ts

@@ -28,7 +28,7 @@ interface ChatActions {
   addMessage: (message: ChatMessage) => void;
   updateMessage: (messageId: number, content: string) => void;
   editMessage: (messageId: number, newContent: string) => Promise<void>;
-  regenerateMessage: (messageId: number) => Promise<void>;
+  regenerateMessage: (messageId: number, model?: string) => Promise<void>;
   feedbackMessage: (messageId: number, feedback: "like" | "dislike") => Promise<void>;
   stopGeneration: () => void;
   
@@ -228,12 +228,6 @@ export const useChatStore = create<ChatStore>((set, get) => ({
     // 创建新的 AbortController
     const abortController = new AbortController();
 
-    console.log("[DEBUG-CHAT] 开始流式发送:", {
-      message: message,
-      sessionId: currentSession.id,
-      knowledgeBaseIds: knowledgeBaseIds
-    });
-
     set({ isStreaming: true, error: null, abortController });
 
     // 添加用户消息
@@ -299,8 +293,7 @@ export const useChatStore = create<ChatStore>((set, get) => ({
               // 处理内容chunk
               chunkCount++;
               totalChars += data.content.length;
-              console.log(`[DEBUG-STREAM] 接收并显示chunk ${chunkCount}:`, data.content);
-              
+
               requestAnimationFrame(() => {
                 set((state) => ({
                   messages: state.messages.map(msg =>
@@ -315,8 +308,7 @@ export const useChatStore = create<ChatStore>((set, get) => ({
             // 向后兼容：纯文本chunk
             chunkCount++;
             totalChars += chunk.length;
-            console.log(`[DEBUG-STREAM] 接收并显示chunk ${chunkCount}:`, chunk);
-            
+
             requestAnimationFrame(() => {
               set((state) => ({
                 messages: state.messages.map(msg =>
@@ -328,17 +320,30 @@ export const useChatStore = create<ChatStore>((set, get) => ({
             });
           }
         },
-        (sessionId: number) => {
-          // 流式完成，直接处理
-          console.log(`[DEBUG-STREAM] 流式完成 - 总chunk数: ${chunkCount}, 总字符数: ${totalChars}`);
-          set({ isStreaming: false, abortController: null });
-          
+        (sessionId: number, messageId?: number, userMessageId?: number) => {
+          if (messageId) {
+            set((state) => ({
+              messages: state.messages.map(msg => {
+                if (msg.id === assistantMessage.id) return { ...msg, id: messageId };
+                if (userMessageId && msg.id === userMessage.id) return { ...msg, id: userMessageId };
+                return msg;
+              }),
+              isStreaming: false,
+              abortController: null,
+            }));
+          } else {
+            const { currentSession } = get();
+            if (currentSession) {
+              get().loadMessages(currentSession.id);
+            }
+            set({ isStreaming: false, abortController: null });
+          }
+
           // 刷新会话列表以确保新会话显示在顶部
           get().loadSessions();
         },
         (error: string) => {
           // 流式错误
-          console.log("[DEBUG-STREAM] 流式错误:", error);
           set({ error, isStreaming: false, abortController: null });
         },
         undefined, // onStatus
@@ -409,20 +414,20 @@ export const useChatStore = create<ChatStore>((set, get) => ({
     }
   },
 
-  // 重新生成消息
-  regenerateMessage: async (messageId: number) => {
+  // 重新生成消息（支持 assistant 消息 ID）
+  regenerateMessage: async (messageId: number, model?: string) => {
     try {
       set({ isLoading: true, error: null });
-      const response = await chatAPI.regenerateMessage(messageId);
-      
-      // 删除该消息之后的所有消息
+      const response = await chatAPI.regenerateMessage(messageId, model);
+
+      // 找到该消息，删除它及之后的所有消息
       const messageIndex = get().messages.findIndex(msg => msg.id === messageId);
       if (messageIndex !== -1) {
         set((state) => ({
-          messages: state.messages.slice(0, messageIndex + 1),
+          messages: state.messages.slice(0, messageIndex),
         }));
       }
-      
+
       // 添加新的助手回复
       const newMessage: ChatMessage = {
         id: response.new_message_id,
@@ -430,7 +435,7 @@ export const useChatStore = create<ChatStore>((set, get) => ({
         content: response.content,
         created_at: new Date().toISOString(),
       };
-      
+
       set((state) => ({
         messages: [...state.messages, newMessage],
         isLoading: false,
@@ -465,9 +470,8 @@ export const useChatStore = create<ChatStore>((set, get) => ({
     // 如果存在 AbortController，调用 abort 中止请求
     if (abortController) {
       abortController.abort();
-      console.log("[DEBUG-STOP] 已中止流式请求");
     }
-    
+
     // 清理状态
     set({ isStreaming: false, abortController: null });
   },