add response versioning support in messages for regeneration improvements

src/routes/chats.js

@@ -105,6 +105,14 @@ const preferredModelValidator = (body) => {
 export function createChatsRouter({repositories, chatService, modelService, config, chatRateLimiters}) {
     const router = express.Router();
     router.use(requireLogin);
+    router.use(["/chats/:id/messages/by-index/:index", "/chats/:id/messages/:messageId"], (req, res, next) => {
+        const chatId = Number(req.params.id);
+        const messageId = req.params.messageId ? Number(req.params.messageId) : chatService.getChatMessageByIndex(chatId, req.params.index)?.id;
+        if (["PUT", "DELETE", "POST"].includes(req.method) && chatService.isRegeneratingMessage(chatId, messageId)) {
+            return res.status(409).json({error: "This response is being regenerated. Wait for it to finish."});
+        }
+        next();
+    });
 
     router.get("/models", async (req, res) => {
         try {
@@ -475,7 +483,21 @@ export function createChatsRouter({repositories, chatService, modelService, conf
         }
     });
 
-    router.post("/chats/:id/messages/:messageId/retry", ...chatRateLimiters, async (req, res) => {
+    router.post("/chats/:id/messages/:messageId/versions/:index", (req, res) => {
+        const chatId = Number(req.params.id), messageId = Number(req.params.messageId);
+        if (!repositories.getChat(chatId, req.session.user)) return res.status(404).json({error: "Chat not found"});
+        const message = repositories.getChatMessage(chatId, messageId);
+        if (!message || message.role !== "bot") return res.status(404).json({error: "Assistant message not found"});
+        if (chatService.isGenerating(chatId)) return res.status(409).json({error: "Wait for generation to finish"});
+        const index = Number(req.params.index);
+        const versions = repositories.listMessageVersions(messageId);
+        if (!Number.isInteger(index) || index < 0 || index >= versions.length) return res.status(400).json({error: "Version is out of range"});
+        repositories.saveMessageVersion(chatId, messageId, null, index);
+        repositories.updateChatContextSummary(chatId, req.session.user, null, 0);
+        res.json({message: chatService.formatChatMessage(repositories.getChatMessage(chatId, messageId))});
+    });
+
+    router.post(["/chats/:id/messages/:messageId/retry", "/chats/:id/messages/:messageId/retry/stream"], ...chatRateLimiters, async (req, res) => {
         const chatId = Number(req.params.id);
         const messageId = Number(req.params.messageId);
         const chat = repositories.getChat(chatId, req.session.user);
@@ -484,19 +506,45 @@ export function createChatsRouter({repositories, chatService, modelService, conf
         if (!targetMessage) return res.status(404).json({error: "Message not found"});
         if (targetMessage.role !== "bot") return res.status(400).json({error: "Only assistant messages can be regenerated"});
 
+        const streaming = req.path.endsWith("/stream");
+        const controller = new AbortController();
+        let finished = false;
+        const sendEvent = (event, data) => {
+            if (!res.destroyed) res.write(`event: ${event}\ndata: ${JSON.stringify(data)}\n\n`);
+        };
+        if (streaming) {
+            res.once("close", () => { if (!finished) controller.abort(); });
+            res.writeHead(200, {"Content-Type": "text/event-stream", "Cache-Control": "no-cache, no-transform", Connection: "keep-alive", "X-Accel-Buffering": "no"});
+        }
         try {
             const result = await chatService.retryMessage({
                 user: req.session.user,
                 chatId,
                 selectedModel: String(req.body?.model || "qwen3:14b"),
                 targetMessage,
-                retryStyle: String(req.body?.style || "").trim() || null
+                retryStyle: String(req.body?.style || "").trim() || null,
+                webSearch: req.body?.webSearch === true,
+                reasoning: req.body?.reasoning === true,
+                signal: controller.signal,
+                ...(streaming ? {
+                    onChunk: (chunk, fullReply) => sendEvent("chunk", {chunk, fullReply}),
+                    onThinking: (chunk, fullThinking) => sendEvent("thinking", {chunk, fullThinking}),
+                    onStatus: status => sendEvent("status", status)
+                } : {})
             });
+            if (streaming) {
+                sendEvent(result.error ? "error" : "done", result.error ? {error: result.error, searchState: result.searchState} : {...result, reply: result.message.content});
+                return;
+            }
             if (result.error) return res.status(result.status).json({error: result.error});
             return res.json(result);
         } catch (error) {
             const mapped = modelService.mapError(error);
+            if (streaming) { sendEvent("error", mapped.body); return; }
             return res.status(mapped.status).json(mapped.body);
+        } finally {
+            finished = true;
+            if (streaming) res.end();
         }
     });
 
@@ -514,7 +562,9 @@ export function createChatsRouter({repositories, chatService, modelService, conf
                 chatId,
                 selectedModel: String(req.body?.model || "qwen3:14b"),
                 targetMessage,
-                retryStyle: String(req.body?.style || "").trim() || null
+                retryStyle: String(req.body?.style || "").trim() || null,
+                webSearch: req.body?.webSearch === true,
+                reasoning: req.body?.reasoning === true
             });
             if (result.error) return res.status(result.status).json({error: result.error});
             return res.json(result);