feat: 企业微信机器人支持接收图片消息自动OCR识别入库

- 注册 message.image 事件监听
- 使用 SDK download_file() 下载并解密图片
- 调用后端 recognize-direct 接口进行 OCR 识别（自动去重）
- 根据识别结果回复：成功录入/重复内容/识别失败
- 更新欢迎语提示支持图片发送功能

app/wework_bot.py

@@ -16,6 +16,9 @@ import logging
 import math
 import os
 import re
+import tempfile
+import time
+from datetime import datetime
 
 import httpx
 from aibot import WSClient, WSClientOptions, generate_req_id
@@ -63,6 +66,7 @@ class WeComBotService:
         # 注册事件
         self.ws_client.on("authenticated", self._on_authenticated)
         self.ws_client.on("message.text", self._on_text_message)
+        self.ws_client.on("message.image", self._on_image_message)
         self.ws_client.on("event.enter_chat", self._on_enter_chat)
         self.ws_client.on("disconnected", self._on_disconnected)
         self.ws_client.on("error", self._on_error)
@@ -83,7 +87,7 @@ class WeComBotService:
             {
                 "msgtype": "text",
                 "text": {
-                    "content": "你好！我是知识库图片搜索助手，发送关键词即可搜索相关图片。"
+                    "content": "你好！我是知识库助手，支持以下功能：\n\n1. 发送关键词 → 搜索相关图片\n2. 发送图片 → 自动识别并录入知识库"
                 },
             },
         )
@@ -194,6 +198,102 @@ class WeComBotService:
         """连接断开回调"""
         logger.warning("企业微信连接断开: %s", reason)
 
+    async def _on_image_message(self, frame: dict):
+        """收到图片消息 -> 下载解密 -> 调用OCR识别 -> 回复结果"""
+        body = frame.get("body", {})
+        image_info = body.get("image", {})
+        image_url = image_info.get("url", "")
+        aes_key = image_info.get("aeskey", "")
+
+        if not image_url:
+            logger.warning("收到图片消息但无下载地址")
+            return
+
+        stream_id = generate_req_id("stream")
+        await self.ws_client.reply_stream(
+            frame, stream_id, "收到图片，正在识别...", False
+        )
+
+        try:
+            # 1. 下载并解密图片
+            image_data, original_filename = await self.ws_client.download_file(
+                image_url, aes_key
+            )
+            if not image_data:
+                await self.ws_client.reply_stream(
+                    frame, stream_id, "图片下载失败，请重试", True
+                )
+                return
+
+            logger.info(
+                "图片下载成功: filename=%s, size=%d bytes",
+                original_filename, len(image_data),
+            )
+
+            # 2. 生成保存文件名（避免重名）
+            ext = os.path.splitext(original_filename or "image.jpg")[1] or ".jpg"
+            timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
+            save_filename = f"wecom_{timestamp}{ext}"
+
+            # 3. 调用后端 recognize-direct 接口
+            async with httpx.AsyncClient(timeout=120.0) as client:
+                resp = await client.post(
+                    f"{self.edubrain_base_url}/api/v1/images/recognize-direct",
+                    files={"file": (save_filename, image_data, "image/jpeg")},
+                )
+
+            if resp.status_code != 200:
+                logger.error("OCR 识别失败: status=%d, body=%s", resp.status_code, resp.text)
+                await self.ws_client.reply_stream(
+                    frame, stream_id, "图片识别失败，请稍后重试", True
+                )
+                return
+
+            result = resp.json()
+
+            # 4. 根据识别结果回复
+            status = result.get("status", "")
+
+            if status == "duplicate":
+                dup_id = result.get("duplicate_of", "?")
+                await self.ws_client.reply_stream(
+                    frame,
+                    stream_id,
+                    f"该图片内容已存在于知识库中（ID: {dup_id}），无需重复录入。",
+                    True,
+                )
+            elif status == "completed":
+                ocr_text = result.get("ocr_text", "")
+                preview = ocr_text[:200].replace("\n", " ") if ocr_text else "无识别内容"
+                record_id = result.get("id", "?")
+                await self.ws_client.reply_stream(
+                    frame,
+                    stream_id,
+                    f"✅ 图片已录入知识库（ID: {record_id}）\n\n识别内容预览：\n{preview}{'...' if len(ocr_text) > 200 else ''}",
+                    True,
+                )
+            elif status == "failed":
+                error_msg = result.get("message", result.get("detail", "未知错误"))
+                await self.ws_client.reply_stream(
+                    frame,
+                    stream_id,
+                    f"图片识别失败：{error_msg}",
+                    True,
+                )
+            else:
+                await self.ws_client.reply_stream(
+                    frame,
+                    stream_id,
+                    f"图片处理结果：{json.dumps(result, ensure_ascii=False)}",
+                    True,
+                )
+
+        except Exception as e:
+            logger.error("处理图片消息失败: %s", e, exc_info=True)
+            await self.ws_client.reply_stream(
+                frame, stream_id, f"处理图片时出错：{e}", True
+            )
+
     def _on_error(self, error):
         """错误回调"""
         logger.error("企业微信机器人错误: %s", error)