Merge pull request #1428 from scut-chenzk/chenzk

修复收到从微信发出的图片消息保存到本地失败的问题

README.md

@@ -27,11 +27,12 @@ Demo made by [Visionn](https://www.wangpc.cc/)
 <img width="240" src="./docs/images/contact.jpg">
 
 # 更新日志
->**2023.09.01：** 增加 [企微个人号](#1385) 通道，[claude](1388)，讯飞星火模型
+
+>**2023.09.01：** 增加 [企微个人号](https://github.com/zhayujie/chatgpt-on-wechat/pull/1385) 通道，[claude](https://github.com/zhayujie/chatgpt-on-wechat/pull/1382) 模型
 
 >**2023.08.08：** 接入百度文心一言模型，通过 [插件](https://github.com/zhayujie/chatgpt-on-wechat/tree/master/plugins/linkai) 支持 Midjourney 绘图
 
->**2023.06.12：** 接入 [LinkAI](https://chat.link-ai.tech/console) 平台，可在线创建个人知识库，并接入微信、公众号及企业微信中，打造专属客服机器人。使用参考 [接入文档](https://link-ai.tech/platform/link-app/wechat)。
+>**2023.06.12：** 接入 [LinkAI](https://chat.link-ai.tech/console) 平台，可在线创建领域知识库，并接入微信、公众号及企业微信中，打造专属客服机器人。使用参考 [接入文档](https://link-ai.tech/platform/link-app/wechat)。
 
 >**2023.04.26：** 支持企业微信应用号部署，兼容插件，并支持语音图片交互，私人助理理想选择，[使用文档](https://github.com/zhayujie/chatgpt-on-wechat/blob/master/channel/wechatcom/README.md)。(contributed by [@lanvent](https://github.com/lanvent) in [#944](https://github.com/zhayujie/chatgpt-on-wechat/pull/944))
 

bot/linkai/link_ai_bot.py

@@ -79,7 +79,10 @@ class LinkAIBot(Bot, OpenAIImage):
                 "frequency_penalty": conf().get("frequency_penalty", 0.0),  # [-2,2]之间，该值越大则更倾向于产生不同的内容
                 "presence_penalty": conf().get("presence_penalty", 0.0),  # [-2,2]之间，该值越大则更倾向于产生不同的内容
             }
-            logger.info(f"[LINKAI] query={query}, app_code={app_code}, mode={body.get('model')}")
+            file_id = context.kwargs.get("file_id")
+            if file_id:
+                body["file_id"] = file_id
+            logger.info(f"[LINKAI] query={query}, app_code={app_code}, mode={body.get('model')}, file_id={file_id}")
             headers = {"Authorization": "Bearer " + linkai_api_key}
 
             # do http request

bridge/bridge.py

@@ -32,6 +32,7 @@ class Bridge(object):
         if model_type in ["claude"]:
             self.btype["chat"] = const.CLAUDEAI
         self.bots = {}
+        self.chat_bots = {}
 
     def get_bot(self, typename):
         if self.bots.get(typename) is None:
@@ -61,6 +62,11 @@ class Bridge(object):
     def fetch_translate(self, text, from_lang="", to_lang="en") -> Reply:
         return self.get_bot("translate").translate(text, from_lang, to_lang)
 
+    def find_chat_bot(self, bot_type: str):
+        if self.chat_bots.get(bot_type) is None:
+            self.chat_bots[bot_type] = create_bot(bot_type)
+        return self.chat_bots.get(bot_type)
+
     def reset_bot(self):
         """
         重置bot路由

bridge/context.py

@@ -9,6 +9,7 @@ class ContextType(Enum):
     IMAGE = 3  # 图片消息
     FILE = 4  # 文件信息
     VIDEO = 5  # 视频信息
+    SHARING = 6  # 分享信息
 
     IMAGE_CREATE = 10  # 创建图片命令
     JOIN_GROUP = 20  # 加入群聊

channel/chat_channel.py

@@ -241,7 +241,7 @@ class ChatChannel(Channel):
                     reply.content = reply_text
                 elif reply.type == ReplyType.ERROR or reply.type == ReplyType.INFO:
                     reply.content = "[" + str(reply.type) + "]\n" + reply.content
-                elif reply.type == ReplyType.IMAGE_URL or reply.type == ReplyType.VOICE or reply.type == ReplyType.IMAGE:
+                elif reply.type == ReplyType.IMAGE_URL or reply.type == ReplyType.VOICE or reply.type == ReplyType.IMAGE or reply.type == ReplyType.FILE or reply.type == ReplyType.VIDEO or reply.type == ReplyType.VIDEO_URL:
                     pass
                 else:
                     logger.error("[WX] unknown reply type: {}".format(reply.type))

channel/wechat/wechat_channel.py

@@ -25,7 +25,7 @@ from lib import itchat
 from lib.itchat.content import *
 
 
-@itchat.msg_register([TEXT, VOICE, PICTURE, NOTE])
+@itchat.msg_register([TEXT, VOICE, PICTURE, NOTE, ATTACHMENT, SHARING])
 def handler_single_msg(msg):
     try:
         cmsg = WechatMessage(msg, False)
@@ -36,7 +36,7 @@ def handler_single_msg(msg):
     return None
 
 
-@itchat.msg_register([TEXT, VOICE, PICTURE, NOTE], isGroupChat=True)
+@itchat.msg_register([TEXT, VOICE, PICTURE, NOTE, ATTACHMENT, SHARING], isGroupChat=True)
 def handler_group_msg(msg):
     try:
         cmsg = WechatMessage(msg, True)
@@ -172,6 +172,8 @@ class WechatChannel(ChatChannel):
         elif cmsg.ctype == ContextType.TEXT:
             # logger.debug("[WX]receive group msg: {}, cmsg={}".format(json.dumps(cmsg._rawmsg, ensure_ascii=False), cmsg))
             pass
+        elif cmsg.ctype == ContextType.FILE:
+            logger.debug(f"[WX]receive attachment msg, file_name={cmsg.content}")
         else:
             logger.debug("[WX]receive group msg: {}".format(cmsg.content))
         context = self._compose_context(cmsg.ctype, cmsg.content, isgroup=True, msg=cmsg)
@@ -208,3 +210,24 @@ class WechatChannel(ChatChannel):
             image_storage.seek(0)
             itchat.send_image(image_storage, toUserName=receiver)
             logger.info("[WX] sendImage, receiver={}".format(receiver))
+        elif reply.type == ReplyType.FILE:  # 新增文件回复类型
+            file_storage = reply.content
+            itchat.send_file(file_storage, toUserName=receiver)
+            logger.info("[WX] sendFile, receiver={}".format(receiver))
+        elif reply.type == ReplyType.VIDEO:  # 新增视频回复类型
+            video_storage = reply.content
+            itchat.send_video(video_storage, toUserName=receiver)
+            logger.info("[WX] sendFile, receiver={}".format(receiver))
+        elif reply.type == ReplyType.VIDEO_URL:  # 新增视频URL回复类型
+            video_url = reply.content
+            logger.debug(f"[WX] start download video, video_url={video_url}")
+            video_res = requests.get(video_url, stream=True)
+            video_storage = io.BytesIO()
+            size = 0
+            for block in video_res.iter_content(1024):
+                size += len(block)
+                video_storage.write(block)
+            logger.info(f"[WX] download video success, size={size}, video_url={video_url}")
+            video_storage.seek(0)
+            itchat.send_video(video_storage, toUserName=receiver)
+            logger.info("[WX] sendVideo url={}, receiver={}".format(video_url, receiver))

channel/wechat/wechat_message.py

@@ -7,7 +7,6 @@ from common.tmp_dir import TmpDir
 from lib import itchat
 from lib.itchat.content import *
 
-
 class WechatMessage(ChatMessage):
     def __init__(self, itchat_msg, is_group=False):
         super().__init__(itchat_msg)
@@ -42,6 +41,14 @@ class WechatMessage(ChatMessage):
                     self.actual_user_nickname = re.findall(r"\"(.*?)\"", itchat_msg["Content"])[0]
             else:
                 raise NotImplementedError("Unsupported note message: " + itchat_msg["Content"])
+        elif itchat_msg["Type"] == ATTACHMENT:
+            self.ctype = ContextType.FILE
+            self.content = TmpDir().path() + itchat_msg["FileName"]
+            self._prepare_fn = lambda: itchat_msg.download(self.content)
+        elif itchat_msg["Type"] == SHARING:
+            self.ctype = ContextType.SHARING
+            self.content = itchat_msg.get("Url")
+
         else:
             raise NotImplementedError("Unsupported message type: Type:{} MsgType:{}".format(itchat_msg["Type"], itchat_msg["MsgType"]))
 

channel/wechatmp/passive_reply.py

@@ -49,7 +49,7 @@ class Query:
 
                 # New request
                 if (
-                    from_user not in channel.cache_dict
+                    channel.cache_dict.get(from_user) is None
                     and from_user not in channel.running
                     or content.startswith("#")
                     and message_id not in channel.request_cnt  # insert the godcmd
@@ -131,8 +131,10 @@ class Query:
 
                 # Only one request can access to the cached data
                 try:
-                    (reply_type, reply_content) = channel.cache_dict.pop(from_user)
-                except KeyError:
+                    (reply_type, reply_content) = channel.cache_dict[from_user].pop(0)
+                    if not channel.cache_dict[from_user]:  # If popping the message makes the list empty, delete the user entry from cache
+                        del channel.cache_dict[from_user]
+                except IndexError:
                     return "success"
 
                 if reply_type == "text":
@@ -146,7 +148,7 @@ class Query:
                             max_split=1,
                         )
                         reply_text = splits[0] + continue_text
-                        channel.cache_dict[from_user] = ("text", splits[1])
+                        channel.cache_dict[from_user].append(("text", splits[1]))
 
                     logger.info(
                         "[wechatmp] Request {} do send to {} {}: {}\n{}".format(

channel/wechatmp/wechatmp_channel.py

@@ -10,6 +10,7 @@ import requests
 import web
 from wechatpy.crypto import WeChatCrypto
 from wechatpy.exceptions import WeChatClientException
+from collections import defaultdict
 
 from bridge.context import *
 from bridge.reply import *
@@ -46,7 +47,7 @@ class WechatMPChannel(ChatChannel):
             self.crypto = WeChatCrypto(token, aes_key, appid)
         if self.passive_reply:
             # Cache the reply to the user's first message
-            self.cache_dict = dict()
+            self.cache_dict = defaultdict(list)
             # Record whether the current message is being processed
             self.running = set()
             # Count the request from wechat official server by message_id
@@ -82,24 +83,28 @@ class WechatMPChannel(ChatChannel):
             if reply.type == ReplyType.TEXT or reply.type == ReplyType.INFO or reply.type == ReplyType.ERROR:
                 reply_text = reply.content
                 logger.info("[wechatmp] text cached, receiver {}\n{}".format(receiver, reply_text))
-                self.cache_dict[receiver] = ("text", reply_text)
+                self.cache_dict[receiver].append(("text", reply_text))
             elif reply.type == ReplyType.VOICE:
-                try:
-                    voice_file_path = reply.content
-                    with open(voice_file_path, "rb") as f:
-                        # support: <2M, <60s, mp3/wma/wav/amr
-                        response = self.client.material.add("voice", f)
-                        logger.debug("[wechatmp] upload voice response: {}".format(response))
-                        # 根据文件大小估计一个微信自动审核的时间，审核结束前返回将会导致语音无法播放，这个估计有待验证
-                        f_size = os.fstat(f.fileno()).st_size
-                        time.sleep(1.0 + 2 * f_size / 1024 / 1024)
-                        # todo check media_id
-                except WeChatClientException as e:
-                    logger.error("[wechatmp] upload voice failed: {}".format(e))
-                    return
-                media_id = response["media_id"]
-                logger.info("[wechatmp] voice uploaded, receiver {}, media_id {}".format(receiver, media_id))
-                self.cache_dict[receiver] = ("voice", media_id)
+                voice_file_path = reply.content
+                duration, files = split_audio(voice_file_path, 60 * 1000)
+                if len(files) > 1:
+                    logger.info("[wechatmp] voice too long {}s > 60s , split into {} parts".format(duration / 1000.0, len(files)))
+
+                for path in files:
+                    # support: <2M, <60s, mp3/wma/wav/amr
+                    try:
+                        with open(path, "rb") as f:
+                            response = self.client.material.add("voice", f)
+                            logger.debug("[wechatmp] upload voice response: {}".format(response))
+                            f_size = os.fstat(f.fileno()).st_size
+                            time.sleep(1.0 + 2 * f_size / 1024 / 1024)
+                            # todo check media_id
+                    except WeChatClientException as e:
+                        logger.error("[wechatmp] upload voice failed: {}".format(e))
+                        return
+                    media_id = response["media_id"]
+                    logger.info("[wechatmp] voice uploaded, receiver {}, media_id {}".format(receiver, media_id))
+                    self.cache_dict[receiver].append(("voice", media_id))
 
             elif reply.type == ReplyType.IMAGE_URL:  # 从网络下载图片
                 img_url = reply.content
@@ -119,7 +124,7 @@ class WechatMPChannel(ChatChannel):
                     return
                 media_id = response["media_id"]
                 logger.info("[wechatmp] image uploaded, receiver {}, media_id {}".format(receiver, media_id))
-                self.cache_dict[receiver] = ("image", media_id)
+                self.cache_dict[receiver].append(("image", media_id))
             elif reply.type == ReplyType.IMAGE:  # 从文件读取图片
                 image_storage = reply.content
                 image_storage.seek(0)
@@ -134,7 +139,7 @@ class WechatMPChannel(ChatChannel):
                     return
                 media_id = response["media_id"]
                 logger.info("[wechatmp] image uploaded, receiver {}, media_id {}".format(receiver, media_id))
-                self.cache_dict[receiver] = ("image", media_id)
+                self.cache_dict[receiver].append(("image", media_id))
         else:
             if reply.type == ReplyType.TEXT or reply.type == ReplyType.INFO or reply.type == ReplyType.ERROR:
                 reply_text = reply.content

channel/wework/wework_message.py

@@ -8,6 +8,7 @@ import pilk
 from bridge.context import ContextType
 from channel.chat_message import ChatMessage
 from common.log import logger
+from ntwork.const import send_type
 
 
 def get_with_retry(get_func, max_retries=5, delay=5):
@@ -41,15 +42,37 @@ def cdn_download(wework, message, file_name):
     data = message["data"]
     aes_key = data["cdn"]["aes_key"]
     file_size = data["cdn"]["size"]
-    file_type = 2
-    file_id = data["cdn"]["file_id"]
 
     # 获取当前工作目录，然后与文件名拼接得到保存路径
     current_dir = os.getcwd()
     save_path = os.path.join(current_dir, "tmp", file_name)
 
-    result = wework.c2c_cdn_download(file_id, aes_key, file_size, file_type, save_path)
-    logger.debug(result)
+    # 下载保存图片到本地
+    if "url" in data["cdn"].keys() and "auth_key" in data["cdn"].keys():
+        url = data["cdn"]["url"]
+        auth_key = data["cdn"]["auth_key"]
+        # result = wework.wx_cdn_download(url, auth_key, aes_key, file_size, save_path)  # ntwork库本身接口有问题，缺失了aes_key这个参数
+        """
+        下载wx类型的cdn文件，以https开头
+        """
+        data = {
+            'url': url,
+            'auth_key': auth_key,
+            'aes_key': aes_key,
+            'size': file_size,
+            'save_path': save_path
+        }
+        result = wework._WeWork__send_sync(send_type.MT_WXCDN_DOWNLOAD_MSG, data)  # 直接用wx_cdn_download的接口内部实现来调用
+    elif "file_id" in data["cdn"].keys():
+        file_type = 2
+        file_id = data["cdn"]["file_id"]
+        result = wework.c2c_cdn_download(file_id, aes_key, file_size, file_type, save_path)
+    else:
+        logger.error(f"something is wrong, data: {data}")
+        return
+
+    # 输出下载结果
+    logger.debug(f"result: {result}")
 
 
 def c2c_download_and_convert(wework, message, file_name):

plugins/godcmd/godcmd.py

@@ -4,8 +4,10 @@ import json
 import os
 import random
 import string
+import logging
 from typing import Tuple
 
+import bridge.bridge
 import plugins
 from bridge.bridge import Bridge
 from bridge.context import ContextType
@@ -310,6 +312,8 @@ class Godcmd(Plugin):
                 elif cmd == "reset":
                     if bottype in [const.OPEN_AI, const.CHATGPT, const.CHATGPTONAZURE, const.LINKAI, const.BAIDU, const.XUNFEI]:
                         bot.sessions.clear_session(session_id)
+                        if Bridge().chat_bots.get(bottype):
+                            Bridge().chat_bots.get(bottype).sessions.clear_session(session_id)
                         channel.cancel_session(session_id)
                         ok, result = True, "会话已重置"
                     else:
@@ -339,8 +343,12 @@ class Godcmd(Plugin):
                             else:
                                 ok, result = False, "当前对话机器人不支持重置会话"
                         elif cmd == "debug":
-                            logger.setLevel("DEBUG")
-                            ok, result = True, "DEBUG模式已开启"
+                            if logger.getEffectiveLevel() == logging.DEBUG:  # 判断当前日志模式是否DEBUG
+                                logger.setLevel(logging.INFO)
+                                ok, result = True, "DEBUG模式已关闭"
+                            else:
+                                logger.setLevel(logging.DEBUG)
+                                ok, result = True, "DEBUG模式已开启"
                         elif cmd == "plist":
                             plugins = PluginManager().list_plugins()
                             ok = True

plugins/keyword/keyword.py

@@ -2,7 +2,7 @@
 
 import json
 import os
-
+import requests
 import plugins
 from bridge.context import ContextType
 from bridge.reply import Reply, ReplyType
@@ -51,15 +51,37 @@ class Keyword(Plugin):
         content = e_context["context"].content.strip()
         logger.debug("[keyword] on_handle_context. content: %s" % content)
         if content in self.keyword:
-            logger.debug(f"[keyword] 匹配到关键字【{content}】")
+            logger.info(f"[keyword] 匹配到关键字【{content}】")
             reply_text = self.keyword[content]
 
             # 判断匹配内容的类型
-            if (reply_text.startswith("http://") or reply_text.startswith("https://")) and any(reply_text.endswith(ext) for ext in [".jpg", ".jpeg", ".png", ".gif", ".webp"]):
-                # 如果是以 http:// 或 https:// 开头，且.jpg/.jpeg/.png/.gif结尾，则认为是图片 URL
+            if (reply_text.startswith("http://") or reply_text.startswith("https://")) and any(reply_text.endswith(ext) for ext in [".jpg", ".jpeg", ".png", ".gif", ".img"]):
+            # 如果是以 http:// 或 https:// 开头，且".jpg", ".jpeg", ".png", ".gif", ".img"结尾，则认为是图片 URL。
                 reply = Reply()
                 reply.type = ReplyType.IMAGE_URL
                 reply.content = reply_text
+                
+            elif (reply_text.startswith("http://") or reply_text.startswith("https://")) and any(reply_text.endswith(ext) for ext in [".pdf", ".doc", ".docx", ".xls", "xlsx",".zip", ".rar"]):
+            # 如果是以 http:// 或 https:// 开头，且".pdf", ".doc", ".docx", ".xls", "xlsx",".zip", ".rar"结尾，则下载文件到tmp目录并发送给用户
+                file_path = "tmp"
+                if not os.path.exists(file_path):
+                    os.makedirs(file_path)
+                file_name = reply_text.split("/")[-1]  # 获取文件名
+                file_path = os.path.join(file_path, file_name)
+                response = requests.get(reply_text)
+                with open(file_path, "wb") as f:
+                    f.write(response.content)
+                #channel/wechat/wechat_channel.py和channel/wechat_channel.py中缺少ReplyType.FILE类型。
+                reply = Reply()
+                reply.type = ReplyType.FILE
+                reply.content = file_path
+            
+            elif (reply_text.startswith("http://") or reply_text.startswith("https://")) and any(reply_text.endswith(ext) for ext in [".mp4"]):
+            # 如果是以 http:// 或 https:// 开头，且".mp4"结尾，则下载视频到tmp目录并发送给用户
+                reply = Reply()
+                reply.type = ReplyType.VIDEO_URL
+                reply.content = reply_text
+                
             else:
             # 否则认为是普通文本
                 reply = Reply()
@@ -68,7 +90,7 @@ class Keyword(Plugin):
             
             e_context["reply"] = reply
             e_context.action = EventAction.BREAK_PASS  # 事件结束，并跳过处理context的默认逻辑
-
+            
     def get_help_text(self, **kwargs):
         help_text = "关键词过滤"
         return help_text

plugins/linkai/README.md

@@ -1,16 +1,16 @@
 ## 插件说明
 
-基于 LinkAI 提供的知识库、Midjourney绘画等能力对机器人的功能进行增强。平台地址: https://chat.link-ai.tech/console
+基于 LinkAI 提供的知识库、Midjourney绘画、文档对话等能力对机器人的功能进行增强。平台地址: https://chat.link-ai.tech/console
 
 ## 插件配置
 
-将 `plugins/linkai` 目录下的 `config.json.template` 配置模板复制为最终生效的 `config.json`。 (如果未配置则会默认使用`config.json.template`模板中配置，功能默认关闭，可通过指令进行开启)。
+将 `plugins/linkai` 目录下的 `config.json.template` 配置模板复制为最终生效的 `config.json`。 (如果未配置则会默认使用`config.json.template`模板中配置，但功能默认关闭，需要可通过指令进行开启)。
 
-以下是配置项说明：
+以下是插件配置项说明：
 
 ```bash
 {
-    "group_app_map": {            # 群聊 和 应用编码 的映射关系
+    "group_app_map": {               # 群聊 和 应用编码 的映射关系
         "测试群名称1": "default",      # 表示在名称为 "测试群名称1" 的群聊中将使用app_code 为 default 的应用
         "测试群名称2": "Kv2fXJcH"
     },
@@ -21,19 +21,30 @@
         "max_tasks": 3,           # 支持同时提交的总任务个数
         "max_tasks_per_user": 1,  # 支持单个用户同时提交的任务个数
         "use_image_create_prefix": true   # 是否使用全局的绘画触发词，如果开启将同时支持由`config.json`中的 image_create_prefix 配置触发
+    },
+    "summary": {
+        "enabled": true,              # 文档总结和对话功能开关
+        "group_enabled": true,        # 是否支持群聊开启
+        "max_file_size": 10000        # 文件的大小限制，单位KB，默认为10M，超过该大小直接忽略
     }
 }
-
 ```
+
+根目录 `config.json` 中配置，`API_KEY` 在 [控制台](https://chat.link-ai.tech/console/interface) 中创建并复制过来:
+
+```bash
+"linkai_api_key": "Link_xxxxxxxxx"
+```
+
 注意：
 
- - 配置项中 `group_app_map` 部分是用于映射群聊与LinkAI平台上的应用， `midjourney` 部分是 mj 画图的配置，可根据需要进行填写，未填写配置时默认不开启相应功能
+ - 配置项中 `group_app_map` 部分是用于映射群聊与LinkAI平台上的应用， `midjourney` 部分是 mj 画图的配置，`summary` 部分是文档总结及对话功能的配置。三部分的配置相互独立，可按需开启
  - 实际 `config.json` 配置中应保证json格式，不应携带 '#' 及后面的注释
  - 如果是`docker`部署，可通过映射 `plugins/config.json` 到容器中来完成插件配置，参考[文档](https://github.com/zhayujie/chatgpt-on-wechat#3-%E6%8F%92%E4%BB%B6%E4%BD%BF%E7%94%A8)
 
 ## 插件使用
 
-> 使用插件中的知识库管理功能需要首先开启`linkai`对话，依赖全局 `config.json` 中的 `use_linkai` 和 `linkai_api_key` 配置；而midjourney绘画功能则只需填写 `linkai_api_key` 配置，`use_linkai` 无论是否关闭均可使用。具体可参考 [详细文档](https://link-ai.tech/platform/link-app/wechat)。
+> 使用插件中的知识库管理功能需要首先开启`linkai`对话，依赖全局 `config.json` 中的 `use_linkai` 和 `linkai_api_key` 配置；而midjourney绘画 和 summary文档总结对话功能则只需填写 `linkai_api_key` 配置，`use_linkai` 无论是否关闭均可使用。具体可参考 [详细文档](https://link-ai.tech/platform/link-app/wechat)。
 
 完成配置后运行项目，会自动运行插件，输入 `#help linkai` 可查看插件功能。
 
@@ -73,7 +84,25 @@
 
 注意事项：
 1. 使用 `$mj open` 和 `$mj close` 指令可以快速打开和关闭绘图功能
-2. 海外环境部署请将 `img_proxy` 设置为 `False`
+2. 海外环境部署请将 `img_proxy` 设置为 `false`
 3. 开启 `use_image_create_prefix` 配置后可直接复用全局画图触发词，以"画"开头便可以生成图片。
 4. 提示词内容中包含敏感词或者参数格式错误可能导致绘画失败，生成失败不消耗积分
 5. 若未收到图片可能有两种可能，一种是收到了图片但微信发送失败，可以在后台日志查看有没有获取到图片url，一般原因是受到了wx限制，可以稍后重试或更换账号尝试；另一种情况是图片提示词存在疑似违规，mj不会直接提示错误但会在画图后删掉原图导致程序无法获取，这种情况不消耗积分。
+
+### 3.文档总结对话功能
+
+#### 配置
+
+该功能依赖 LinkAI的知识库及对话功能，需要在项目根目录的config.json中设置 `linkai_api_key`， 同时根据上述插件配置说明，在插件config.json添加 `summary` 部分的配置，设置 `enabled` 为 true。
+
+如果不想创建 `plugins/linkai/config.json` 配置，可以直接通过 `$linkai sum open` 指令开启该功能。
+
+#### 使用
+
+功能开启后，向机器人发送 **文件** 或 **分享链接卡片** 即可生成摘要，进一步可以与文件或链接的内容进行多轮对话。
+
+#### 限制
+
+ 1. 文件目前 支持 `txt`, `docx`, `pdf`, `md`, `csv`格式，文件大小由 `max_file_size` 限制，最大不超过15M，文件字数最多可支持百万字的文件。但不建议上传字数过多的文件，一是token消耗过大，二是摘要很难覆盖到全部内容，只能通过多轮对话来了解细节。
+ 2. 分享链接 目前仅支持 公众号文章，后续会支持更多文章类型及视频链接等
+ 3. 总结及对话的 费用与 LinkAI 3.5-4K 模型的计费方式相同，按文档内容的tokens进行计算

plugins/linkai/config.json.template

@@ -10,5 +10,10 @@
         "max_tasks": 3,
         "max_tasks_per_user": 1,
         "use_image_create_prefix": true
+    },
+    "summary": {
+        "enabled": true,
+        "group_enabled": true,
+        "max_file_size": 15000
     }
 }

plugins/linkai/linkai.py

@@ -4,7 +4,11 @@ from bridge.reply import Reply, ReplyType
 from config import global_config
 from plugins import *
 from .midjourney import MJBot
+from .summary import LinkSummary
 from bridge import bridge
+from common.expired_dict import ExpiredDict
+from common import const
+import os
 
 
 @plugins.register(
@@ -12,6 +16,7 @@ from bridge import bridge
     desc="A plugin that supports knowledge base and midjourney drawing.",
     version="0.1.0",
     author="https://link-ai.tech",
+    desire_priority=99
 )
 class LinkAI(Plugin):
     def __init__(self):
@@ -23,8 +28,12 @@ class LinkAI(Plugin):
             self.config = self._load_config_template()
         if self.config:
             self.mj_bot = MJBot(self.config.get("midjourney"))
+        self.sum_config = {}
+        if self.config:
+            self.sum_config = self.config.get("summary")
         logger.info("[LinkAI] inited")
 
+
     def on_handle_context(self, e_context: EventContext):
         """
         消息处理逻辑
@@ -34,10 +43,39 @@ class LinkAI(Plugin):
             return
 
         context = e_context['context']
-        if context.type not in [ContextType.TEXT, ContextType.IMAGE, ContextType.IMAGE_CREATE]:
+        if context.type not in [ContextType.TEXT, ContextType.IMAGE, ContextType.IMAGE_CREATE, ContextType.FILE, ContextType.SHARING]:
             # filter content no need solve
             return
 
+        if context.type == ContextType.FILE and self._is_summary_open(context):
+            # 文件处理
+            context.get("msg").prepare()
+            file_path = context.content
+            if not LinkSummary().check_file(file_path, self.sum_config):
+                return
+            _send_info(e_context, "正在为你加速生成摘要，请稍后")
+            res = LinkSummary().summary_file(file_path)
+            if not res:
+                _set_reply_text("总结出现异常，请稍后再试吧", e_context)
+                return
+            USER_FILE_MAP[_find_user_id(context) + "-sum_id"] = res.get("summary_id")
+            _set_reply_text(res.get("summary") + "\n\n💬 发送 \"开启对话\" 可以开启与文件内容的对话", e_context, level=ReplyType.TEXT)
+            os.remove(file_path)
+            return
+
+        if (context.type == ContextType.SHARING and self._is_summary_open(context)) or \
+                (context.type == ContextType.TEXT and LinkSummary().check_url(context.content)):
+            if not LinkSummary().check_url(context.content):
+                return
+            _send_info(e_context, "正在为你加速生成摘要，请稍后")
+            res = LinkSummary().summary_url(context.content)
+            if not res:
+                _set_reply_text("总结出现异常，请稍后再试吧", e_context)
+                return
+            _set_reply_text(res.get("summary") + "\n\n💬 发送 \"开启对话\" 可以开启与文章内容的对话", e_context, level=ReplyType.TEXT)
+            USER_FILE_MAP[_find_user_id(context) + "-sum_id"] = res.get("summary_id")
+            return
+
         mj_type = self.mj_bot.judge_mj_task_type(e_context)
         if mj_type:
             # MJ作图任务处理
@@ -49,10 +87,38 @@ class LinkAI(Plugin):
             self._process_admin_cmd(e_context)
             return
 
+        if context.type == ContextType.TEXT and context.content == "开启对话" and _find_sum_id(context):
+            # 文本对话
+            _send_info(e_context, "正在为你开启对话，请稍后")
+            res = LinkSummary().summary_chat(_find_sum_id(context))
+            if not res:
+                _set_reply_text("开启对话失败，请稍后再试吧", e_context)
+                return
+            USER_FILE_MAP[_find_user_id(context) + "-file_id"] = res.get("file_id")
+            _set_reply_text("💡你可以问我关于这篇文章的任何问题，例如：\n\n" + res.get("questions") + "\n\n发送 \"退出对话\" 可以关闭与文章的对话", e_context, level=ReplyType.TEXT)
+            return
+
+        if context.type == ContextType.TEXT and context.content == "退出对话" and _find_file_id(context):
+            del USER_FILE_MAP[_find_user_id(context) + "-file_id"]
+            bot = bridge.Bridge().find_chat_bot(const.LINKAI)
+            bot.sessions.clear_session(context["session_id"])
+            _set_reply_text("对话已退出", e_context, level=ReplyType.TEXT)
+            return
+
+        if context.type == ContextType.TEXT and _find_file_id(context):
+            bot = bridge.Bridge().find_chat_bot(const.LINKAI)
+            context.kwargs["file_id"] = _find_file_id(context)
+            reply = bot.reply(context.content, context)
+            e_context["reply"] = reply
+            e_context.action = EventAction.BREAK_PASS
+            return
+
+
         if self._is_chat_task(e_context):
             # 文本对话任务处理
             self._process_chat_task(e_context)
 
+
     # 插件管理功能
     def _process_admin_cmd(self, e_context: EventContext):
         context = e_context['context']
@@ -94,11 +160,36 @@ class LinkAI(Plugin):
             # 保存插件配置
             super().save_config(self.config)
             _set_reply_text(f"应用设置成功: {app_code}", e_context, level=ReplyType.INFO)
-        else:
-            _set_reply_text(f"指令错误，请输入{_get_trigger_prefix()}linkai help 获取帮助", e_context,
-                            level=ReplyType.INFO)
             return
 
+        if len(cmd) == 3 and cmd[1] == "sum" and (cmd[2] == "open" or cmd[2] == "close"):
+            # 知识库开关指令
+            if not _is_admin(e_context):
+                _set_reply_text("需要管理员权限执行", e_context, level=ReplyType.ERROR)
+                return
+            is_open = True
+            tips_text = "开启"
+            if cmd[2] == "close":
+                tips_text = "关闭"
+                is_open = False
+            if not self.sum_config:
+                _set_reply_text(f"插件未启用summary功能，请参考以下链添加插件配置\n\nhttps://github.com/zhayujie/chatgpt-on-wechat/blob/master/plugins/linkai/README.md", e_context, level=ReplyType.INFO)
+            else:
+                self.sum_config["enabled"] = is_open
+                _set_reply_text(f"文章总结功能{tips_text}", e_context, level=ReplyType.INFO)
+            return
+
+        _set_reply_text(f"指令错误，请输入{_get_trigger_prefix()}linkai help 获取帮助", e_context,
+                        level=ReplyType.INFO)
+        return
+
+    def _is_summary_open(self, context) -> bool:
+        if not self.sum_config or not self.sum_config.get("enabled"):
+            return False
+        if not context.kwargs.get("isgroup") and not self.sum_config.get("group_enabled"):
+            return False
+        return True
+
     # LinkAI 对话任务处理
     def _is_chat_task(self, e_context: EventContext):
         context = e_context['context']
@@ -112,7 +203,7 @@ class LinkAI(Plugin):
         """
         context = e_context['context']
         # 群聊应用管理
-        group_name = context.kwargs.get("msg").from_user_nickname
+        group_name = context.get("msg").from_user_nickname
         app_code = self._fetch_group_app_code(group_name)
         if app_code:
             context.kwargs['app_code'] = app_code
@@ -130,7 +221,7 @@ class LinkAI(Plugin):
 
     def get_help_text(self, verbose=False, **kwargs):
         trigger_prefix = _get_trigger_prefix()
-        help_text = "用于集成 LinkAI 提供的知识库、Midjourney绘画等能力。\n\n"
+        help_text = "用于集成 LinkAI 提供的知识库、Midjourney绘画、文档总结对话等能力。\n\n"
         if not verbose:
             return help_text
         help_text += f'📖 知识库\n - 群聊中指定应用: {trigger_prefix}linkai app 应用编码\n'
@@ -140,6 +231,7 @@ class LinkAI(Plugin):
         help_text += f"🎨 绘画\n - 生成: {trigger_prefix}mj 描述词1, 描述词2.. \n - 放大: {trigger_prefix}mju 图片ID 图片序号\n - 变换: {trigger_prefix}mjv 图片ID 图片序号\n - 重置: {trigger_prefix}mjr 图片ID"
         help_text += f"\n\n例如：\n\"{trigger_prefix}mj a little cat, white --ar 9:16\"\n\"{trigger_prefix}mju 11055927171882 2\""
         help_text += f"\n\"{trigger_prefix}mjv 11055927171882 2\"\n\"{trigger_prefix}mjr 11055927171882\""
+        help_text += f"\n\n💡 文档总结和对话\n - 开启: {trigger_prefix}linkai sum open\n - 使用: 发送文件、公众号文章等可生成摘要，并与内容对话"
         return help_text
 
     def _load_config_template(self):
@@ -150,10 +242,17 @@ class LinkAI(Plugin):
                 with open(plugin_config_path, "r", encoding="utf-8") as f:
                     plugin_conf = json.load(f)
                     plugin_conf["midjourney"]["enabled"] = False
+                    plugin_conf["summary"]["enabled"] = False
                     return plugin_conf
         except Exception as e:
             logger.exception(e)
 
+
+def _send_info(e_context: EventContext, content: str):
+    reply = Reply(ReplyType.TEXT, content)
+    channel = e_context["channel"]
+    channel.send(reply, e_context["context"])
+
 # 静态方法
 def _is_admin(e_context: EventContext) -> bool:
     """
@@ -168,11 +267,26 @@ def _is_admin(e_context: EventContext) -> bool:
         return context["receiver"] in global_config["admin_users"]
 
 
+def _find_user_id(context):
+    if context["isgroup"]:
+        return context.kwargs.get("msg").actual_user_id
+    else:
+        return context["receiver"]
+
+
 def _set_reply_text(content: str, e_context: EventContext, level: ReplyType = ReplyType.ERROR):
     reply = Reply(level, content)
     e_context["reply"] = reply
     e_context.action = EventAction.BREAK_PASS
 
-
 def _get_trigger_prefix():
     return conf().get("plugin_trigger_prefix", "$")
+
+def _find_sum_id(context):
+    return USER_FILE_MAP.get(_find_user_id(context) + "-sum_id")
+
+def _find_file_id(context):
+    return USER_FILE_MAP.get(_find_user_id(context) + "-file_id")
+
+
+USER_FILE_MAP = ExpiredDict(conf().get("expires_in_seconds") or 60 * 60)

plugins/linkai/midjourney.py

@@ -5,7 +5,6 @@ import requests
 import threading
 import time
 from bridge.reply import Reply, ReplyType
-import aiohttp
 import asyncio
 from bridge.context import ContextType
 from plugins import EventContext, EventAction

plugins/linkai/summary.py

@@ -0,0 +1,89 @@
+import requests
+from config import conf
+from common.log import logger
+import os
+
+
+class LinkSummary:
+    def __init__(self):
+        pass
+
+    def summary_file(self, file_path: str):
+        file_body = {
+            "file": open(file_path, "rb"),
+            "name": file_path.split("/")[-1],
+        }
+        res = requests.post(url=self.base_url() + "/v1/summary/file", headers=self.headers(), files=file_body, timeout=(5, 180))
+        return self._parse_summary_res(res)
+
+    def summary_url(self, url: str):
+        body = {
+            "url": url
+        }
+        res = requests.post(url=self.base_url() + "/v1/summary/url", headers=self.headers(), json=body, timeout=(5, 180))
+        return self._parse_summary_res(res)
+
+    def summary_chat(self, summary_id: str):
+        body = {
+            "summary_id": summary_id
+        }
+        res = requests.post(url=self.base_url() + "/v1/summary/chat", headers=self.headers(), json=body, timeout=(5, 180))
+        if res.status_code == 200:
+            res = res.json()
+            logger.debug(f"[LinkSum] chat open, res={res}")
+            if res.get("code") == 200:
+                data = res.get("data")
+                return {
+                    "questions": data.get("questions"),
+                    "file_id": data.get("file_id")
+                }
+        else:
+            res_json = res.json()
+            logger.error(f"[LinkSum] summary error, status_code={res.status_code}, msg={res_json.get('message')}")
+            return None
+
+    def _parse_summary_res(self, res):
+        if res.status_code == 200:
+            res = res.json()
+            logger.debug(f"[LinkSum] url summary, res={res}")
+            if res.get("code") == 200:
+                data = res.get("data")
+                return {
+                    "summary": data.get("summary"),
+                    "summary_id": data.get("summary_id")
+                }
+        else:
+            res_json = res.json()
+            logger.error(f"[LinkSum] summary error, status_code={res.status_code}, msg={res_json.get('message')}")
+            return None
+
+    def base_url(self):
+        return conf().get("linkai_api_base", "https://api.link-ai.chat")
+
+    def headers(self):
+        return {"Authorization": "Bearer " + conf().get("linkai_api_key")}
+
+    def check_file(self, file_path: str, sum_config: dict) -> bool:
+        file_size = os.path.getsize(file_path) // 1000
+
+        if (sum_config.get("max_file_size") and file_size > sum_config.get("max_file_size")) or file_size > 15000:
+            logger.warn(f"[LinkSum] file size exceeds limit, No processing, file_size={file_size}KB")
+            return True
+
+        suffix = file_path.split(".")[-1]
+        support_list = ["txt", "csv", "docx", "pdf", "md"]
+        if suffix not in support_list:
+            logger.warn(f"[LinkSum] unsupported file, suffix={suffix}, support_list={support_list}")
+            return False
+
+        return True
+
+    def check_url(self, url: str):
+        if not url:
+            return False
+        support_list = ["http://mp.weixin.qq.com", "https://mp.weixin.qq.com"]
+        for support_url in support_list:
+            if url.strip().startswith(support_url):
+                return True
+        logger.debug("[LinkSum] unsupported url, no need to process")
+        return False