语音回复后增加文本回复（目前在公众号和企业微信应用支持）

app.py

@@ -11,15 +11,19 @@
 from plugins import *
 import threading
 
+shutdown = False
 
 def sigterm_handler_wrap(_signo):
     old_handler = signal.getsignal(_signo)
 
     def func(_signo, _stack_frame):
+        global shutdown
         logger.info("signal {} received, exiting...".format(_signo))
         conf().save_user_datas()
-        if callable(old_handler):  #  check old_handler
-            return old_handler(_signo, _stack_frame)
+
+        # if callable(old_handler):  # check old_handler
+        #     return old_handler(_signo, _stack_frame)
+        shutdown = True
         sys.exit(0)
 
     signal.signal(_signo, func)
@@ -60,8 +64,10 @@ def run():
 
         start_channel(channel_name)
 
-        while True:
+        while not shutdown:
             time.sleep(1)
+
+        logger.info("exited in run()")
     except Exception as e:
         logger.error("App startup failed!")
         logger.exception(e)

bot/zhipuai/zhipu_ai_session.py

@@ -5,9 +5,10 @@
 class ZhipuAISession(Session):
     def __init__(self, session_id, system_prompt=None, model="glm-4"):
         super().__init__(session_id, system_prompt)
+
         self.model = model
         self.reset()
-        if not system_prompt:
+        if not self.system_prompt:
             logger.warn("[ZhiPu] `character_desc` can not be empty")
 
     def discard_exceeding(self, max_tokens, cur_tokens=None):

bridge/bridge.py

@@ -34,7 +34,7 @@ def __init__(self):
             self.btype["chat"] = const.QWEN_DASHSCOPE
         if model_type in [const.GEMINI]:
             self.btype["chat"] = const.GEMINI
-        if model_type in [const.ZHIPU_AI]:
+        if model_type in [const.ZHIPU_AI,const.ZHIPU_AI_GLM4_FLASH,const.ZHIPU_AI_GLM4_AIR]:
             self.btype["chat"] = const.ZHIPU_AI
         if model_type and model_type.startswith("claude-3"):
             self.btype["chat"] = const.CLAUDEAPI

bridge/context.py

@@ -17,7 +17,7 @@ class ContextType(Enum):
     PATPAT = 21  # 拍了拍
     FUNCTION = 22  # 函数调用
     EXIT_GROUP = 23 #退出
-
+    RETELL = 30  # 文本朗读命令
 
     def __str__(self):
         return self.name

bridge/reply.py

@@ -26,6 +26,7 @@ class Reply:
     def __init__(self, type: ReplyType = None, content=None):
         self.type = type
         self.content = content
+        self.orig_content = None
 
     def __str__(self):
-        return "Reply(type={}, content={})".format(self.type, self.content)
+        return "Reply(type={}, content={}, orig_content={})".format(self.type, self.content, self.orig_content)

channel/chat_channel.py

@@ -152,7 +152,14 @@ def _compose_context(self, ctype: ContextType, content, **kwargs):
                 content = content.replace(img_match_prefix, "", 1)
                 context.type = ContextType.IMAGE_CREATE
             else:
-                context.type = ContextType.TEXT
+                retell_match_prefix = check_prefix(content, conf().get("retell_prefix"))
+                if retell_match_prefix:
+                    content = content.replace(retell_match_prefix, "", 1)
+                    context.type = ContextType.RETELL
+                    context["desire_rtype"] = ReplyType.VOICE
+                else:
+                    context.type = ContextType.TEXT
+
             context.content = content.strip()
             if "desire_rtype" not in context and conf().get("always_reply_voice") and ReplyType.VOICE not in self.NOT_SUPPORT_REPLYTYPE:
                 context["desire_rtype"] = ReplyType.VOICE
@@ -190,6 +197,8 @@ def _generate_reply(self, context: Context, reply: Reply = Reply()) -> Reply:
             if context.type == ContextType.TEXT or context.type == ContextType.IMAGE_CREATE:  # 文字和图片消息
                 context["channel"] = e_context["channel"]
                 reply = super().build_reply_content(context.content, context)
+            elif context.type == ContextType.RETELL:  # 朗读命令
+                reply = super().build_text_to_voice(context.content)
             elif context.type == ContextType.VOICE:  # 语音消息
                 cmsg = context["msg"]
                 cmsg.prepare()
@@ -251,6 +260,7 @@ def _decorate_reply(self, context: Context, reply: Reply) -> Reply:
                     reply_text = reply.content
                     if desire_rtype == ReplyType.VOICE and ReplyType.VOICE not in self.NOT_SUPPORT_REPLYTYPE:
                         reply = super().build_text_to_voice(reply.content)
+                        reply.orig_content = reply_text
                         return self._decorate_reply(context, reply)
                     if context.get("isgroup", False):
                         if not context.get("no_need_at", False):

channel/wechatcom/wechatcomapp_channel.py

@@ -21,6 +21,11 @@
 from config import conf, subscribe_msg
 from voice.audio_convert import any_to_amr, split_audio
 
+from wechatpy import events
+from wechatpy.fields import IntegerField, StringField
+import xmltodict
+from wechatpy.utils import to_text
+
 MAX_UTF8_LEN = 2048
 
 
@@ -41,23 +46,32 @@ def __init__(self):
         )
         self.crypto = WeChatCrypto(self.token, self.aes_key, self.corp_id)
         self.client = WechatComAppClient(self.corp_id, self.secret)
+        self.text_after_voice = conf().get("text_after_voice", False)
 
     def startup(self):
         # start message listener
-        urls = ("/wxcomapp", "channel.wechatcom.wechatcomapp_channel.Query")
+        urls = (conf().get("wechatcomapp_url", "/wxcomapp"), "channel.wechatcom.wechatcomapp_channel.Query")
         app = web.application(urls, globals(), autoreload=False)
         port = conf().get("wechatcomapp_port", 9898)
         web.httpserver.runsimple(app.wsgifunc(), ("0.0.0.0", port))
 
     def send(self, reply: Reply, context: Context):
         receiver = context["receiver"]
+        logger.debug("[wechatcom] context {} ".format(context.kwargs['msg']))
+        if context.kf_mode:
+            receiver = context.kwargs['msg'].from_user_id  # 客服模式下，external_userid 就是客户id
+            agent_id = context.kwargs['msg'].to_user_id  # 客服模式下，agent_id 就是客服id
+        else:
+            agent_id = self.agent_id  # 非客服模式下，agent_id 就是应用的 agent_id
+
         if reply.type in [ReplyType.TEXT, ReplyType.ERROR, ReplyType.INFO]:
             reply_text = reply.content
             texts = split_string_by_utf8_length(reply_text, MAX_UTF8_LEN)
             if len(texts) > 1:
                 logger.info("[wechatcom] text too long, split into {} parts".format(len(texts)))
             for i, text in enumerate(texts):
-                self.client.message.send_text(self.agent_id, receiver, text)
+                self.send_text_message(agent_id, receiver, text, context.kf_mode)
+
                 if i != len(texts) - 1:
                     time.sleep(0.5)  # 休眠0.5秒，防止发送过快乱序
             logger.info("[wechatcom] Do send text to {}: {}".format(receiver, reply_text))
@@ -84,9 +98,15 @@ def send(self, reply: Reply, context: Context):
             except Exception:
                 pass
             for media_id in media_ids:
-                self.client.message.send_voice(self.agent_id, receiver, media_id)
+                self.send_voice_message(agent_id, receiver, media_id, context.kf_mode)
                 time.sleep(1)
             logger.info("[wechatcom] sendVoice={}, receiver={}".format(reply.content, receiver))
+
+            # if need text_after_voice
+            if self.text_after_voice and reply.orig_content:
+                logger.debug("[wechatcom] send text after voice: {}".format(reply.orig_content))
+                self.send_text_message(agent_id, receiver, reply.orig_content, context.kf_mode)
+
         elif reply.type == ReplyType.IMAGE_URL:  # 从网络下载图片
             img_url = reply.content
             pic_res = requests.get(img_url, stream=True)
@@ -106,7 +126,7 @@ def send(self, reply: Reply, context: Context):
                 logger.error("[wechatcom] upload image failed: {}".format(e))
                 return
 
-            self.client.message.send_image(self.agent_id, receiver, response["media_id"])
+            self.send_image_message(agent_id, receiver, response["media_id"], context.kf_mode)
             logger.info("[wechatcom] sendImage url={}, receiver={}".format(img_url, receiver))
         elif reply.type == ReplyType.IMAGE:  # 从文件读取图片
             image_storage = reply.content
@@ -122,9 +142,99 @@ def send(self, reply: Reply, context: Context):
             except WeChatClientException as e:
                 logger.error("[wechatcom] upload image failed: {}".format(e))
                 return
-            self.client.message.send_image(self.agent_id, receiver, response["media_id"])
+            self.send_image_message(agent_id, receiver, response["media_id"], context.kf_mode)
             logger.info("[wechatcom] sendImage, receiver={}".format(receiver))
 
+    def send_text_message(self, agent_id, receiver, content, kf_mode):
+        if not kf_mode:
+            return self.client.message.send_text(agent_id, receiver, content)
+
+        url = f"https://qyapi.weixin.qq.com/cgi-bin/kf/send_msg?access_token={self.client.fetch_access_token_cs()}"
+        data = {
+            "touser": receiver,
+            "open_kfid": agent_id,
+            "msgtype": "text",
+            "text": {"content": content}
+        }
+
+        response = requests.post(url, json=data)
+        return response.json()
+
+    def send_image_message(self, agent_id, receiver, media_id, kf_mode):
+        if not kf_mode:
+            return self.client.message.send_image(agent_id, receiver, media_id)
+
+        url = f"https://qyapi.weixin.qq.com/cgi-bin/kf/send_msg?access_token={self.client.fetch_access_token_cs()}"
+        data = {
+            "touser": receiver,
+            "open_kfid": agent_id,
+            "msgtype": "image",
+            "image": {"media_id": media_id}
+        }
+
+        response = requests.post(url, json=data).json()
+        if response['errmsg'] == 'ok':
+            logger.debug(f"Send IMAGE Message Success")
+        else:
+            logger.error(f"Something error:{response}")
+        return response
+
+    def send_voice_message(self, agent_id, receiver, media_id, kf_mode):
+        if not kf_mode:
+            return self.client.message.send_voice(agent_id, receiver, media_id)
+
+        url = f"https://qyapi.weixin.qq.com/cgi-bin/kf/send_msg?access_token={self.client.fetch_access_token_cs()}"
+        data = {
+            "touser": receiver,
+            "open_kfid": agent_id,
+            "msgtype": "voice",
+            "voice": {"media_id": media_id}
+        }
+
+        response = requests.post(url, json=data).json()
+        if response['errmsg'] == 'ok':
+            logger.debug(f"Send VOICE Message Success")
+        else:
+            logger.error(f"Something error:{response}")
+        return response
+
+    def get_latest_message(self, token, open_kfid, next_cursor=""):
+        url = f"https://qyapi.weixin.qq.com/cgi-bin/kf/sync_msg?access_token={self.client.fetch_access_token_cs()}"
+        data = {
+            "token": token,
+            "open_kfid": open_kfid,
+            "limit": 1000
+        }
+        if next_cursor:
+            data["cursor"] = next_cursor
+
+        response = requests.post(url, json=data)
+        response_data = response.json()
+
+        # 检查是否有错误码并打印相关错误信息
+        if response_data.get("errcode") != 0:
+            logger.error(
+                f"[ERROR][{response_data.get('errcode')}][{response_data.get('errmsg')}] - Failed to fetch messages, more info at {response_data.get('more_info') or 'https://open.work.weixin.qq.com/devtool/query?e=' + str(response_data.get('errcode'))}")
+            return None
+
+        # logger.debug(f"response_data:{response_data}")
+        if response_data.get("msg_list"):
+            return response_data["msg_list"][-1]  # 返回最新的一条消息
+        else:
+            return None
+
+# 扩展客服消息事件类
+class CustomServiceEvent(events.BaseEvent):
+    """
+    客服消息或事件
+    """
+    agent = IntegerField('AgentID', 0)
+    event = 'kf_msg_or_event'
+    source = StringField('FromUserName')
+    target = StringField('ToUserName')
+    time = IntegerField('CreateTime')
+    token = StringField('Token')
+    open_kfid = StringField('OpenKfId')
 
 class Query:
     def GET(self):
@@ -138,6 +248,7 @@ def GET(self):
             echostr = params.echostr
             echostr = channel.crypto.check_signature(signature, timestamp, nonce, echostr)
         except InvalidSignatureException:
+            logger.error("[wechatcom] Invalid signature in GET request")
             raise web.Forbidden()
         return echostr
 
@@ -150,29 +261,56 @@ def POST(self):
             timestamp = params.timestamp
             nonce = params.nonce
             message = channel.crypto.decrypt_message(web.data(), signature, timestamp, nonce)
+            msg = self.extended_parse_message(message)
+            logger.debug("[wechatcom] receive message: {}, msg= {}".format(message, msg))
+
         except (InvalidSignatureException, InvalidCorpIdException):
             raise web.Forbidden()
-        msg = parse_message(message)
-        logger.debug("[wechatcom] receive message: {}, msg= {}".format(message, msg))
+
+        kf_msg = None
         if msg.type == "event":
             if msg.event == "subscribe":
                 reply_content = subscribe_msg()
                 if reply_content:
                     reply = create_reply(reply_content, msg).render()
                     res = channel.crypto.encrypt_message(reply, nonce, timestamp)
                     return res
-        else:
-            try:
-                wechatcom_msg = WechatComAppMessage(msg, client=channel.client)
-            except NotImplementedError as e:
-                logger.debug("[wechatcom] " + str(e))
+            elif msg.event == "kf_msg_or_event":
+                kf_msg = channel.get_latest_message(msg.token, msg.open_kfid)
+                logger.debug("[wechatcom] latest_message: {}".format(msg))
+            else:
+                logger.debug("[wechatcom] receive unsupported event: {}".format(msg.event))
                 return "success"
-            context = channel._compose_context(
-                wechatcom_msg.ctype,
-                wechatcom_msg.content,
-                isgroup=False,
-                msg=wechatcom_msg,
-            )
-            if context:
-                channel.produce(context)
+
+        try:
+            wechatcom_msg = WechatComAppMessage(msg, client=channel.client, kf_msg=kf_msg)
+        except NotImplementedError as e:
+            logger.debug("[wechatcom] " + str(e))
+            return "success"
+        context = channel._compose_context(
+            wechatcom_msg.ctype,
+            wechatcom_msg.content,
+            isgroup=False,
+            msg=wechatcom_msg,
+        )
+        if context:
+            context.kf_mode = kf_msg is not None    #是否客服模式
+            channel.produce(context)
         return "success"
+
+
+    # 扩展解析消息
+    def extended_parse_message(self, message):
+        msg = parse_message(message)
+        if msg.type != "unknown":
+            return msg
+
+        # 尝试解析客服消息事件
+        msg_text = xmltodict.parse(to_text(message))['xml']
+        message_type = msg_text['MsgType'].lower()
+        if message_type == 'event':
+            event_type = msg_text['Event'].lower()
+            if event_type == "kf_msg_or_event":
+                return CustomServiceEvent(msg_text)
+
+        return msg

channel/wechatcom/wechatcomapp_client.py

@@ -1,8 +1,8 @@
 import threading
 import time
-
+import requests
 from wechatpy.enterprise import WeChatClient
-
+from config import conf
 
 class WechatComAppClient(WeChatClient):
     def __init__(self, corp_id, secret, access_token=None, session=None, timeout=None, auto_retry=True):
@@ -19,3 +19,21 @@ def fetch_access_token(self):  # 重载父类方法，加锁避免多线程重
                 if self.expires_at - timestamp > 60:
                     return access_token
             return super().fetch_access_token()
+
+    def fetch_access_token_cs(self):
+        current_time = time.time()
+        if self.access_token and self.expires_at - current_time > 60:
+            return self.access_token
+
+        corpid = conf().get("wechatcom_corp_id")
+        corpsecret = conf().get("wechatcomapp_secret")
+        url = f"https://qyapi.weixin.qq.com/cgi-bin/gettoken?corpid={corpid}&corpsecret={corpsecret}"
+
+        response = requests.get(url).json()
+        if 'access_token' in response:
+            self.access_token = response['access_token']
+            self.expires_at = current_time + response['expires_in'] - 60
+            print(f'access_token:{self.access_token}')
+            return self.access_token
+        else:
+            raise Exception("Failed to retrieve access token")