J1147
/
ai-on-wechat
Sledovat 1
Oblíbit 0
Rozštěpit 0
Zdrojový kód Úkoly 0 Požadavky na natažení 0 Vydání 0 Wiki Aktivita
Nelze vybrat více než 25 témat Téma musí začínat písmenem nebo číslem, může obsahovat pomlčky („-“) a může být dlouhé až 35 znaků.
1202 Revize
2 Větve
152MB
Strom: 36981bf516
Větve Značky
${ item.name }
Vytvořit větev ${ searchTerm }
z „36981bf516“
${ noResults }
ai-on-wechat/plugins/coze4upload/coze4upload.py

372 lines
16KB
Surový Blame Historie 

  1. import requests

  2. import json

  3. import plugins

  4. from bridge.reply import Reply, ReplyType

  5. from bridge.context import ContextType

  6. from channel.chat_message import ChatMessage

  7. from plugins import *

  8. from common.log import logger

  9. from common.expired_dict import ExpiredDict

  10. import os

  11. import base64

  12. from pathlib import Path

  13. from PIL import Image

  14. import oss2

  15. from lib import itchat

  16. from lib.itchat.content import *

  17. import re

  18. 

  19. 

  20. # C:\Users\vsoni\source\repos\chatgpt-on-wechat\channel\wechat\wechat_channel.py

  21. 

  22. @plugins.register(

  23.     name="coze4upload",

  24.     desire_priority=-1,

  25.     desc="A plugin for upload",

  26.     version="0.0.01",

  27.     author="",

  28. )

  29. 

  30. class coze4upload(Plugin):

  31.     def __init__(self):

  32.         super().__init__()

  33.         try:

  34.             curdir = os.path.dirname(__file__)

  35.             config_path = os.path.join(curdir, "config.json")

  36.             if os.path.exists(config_path):

  37.                 with open(config_path, "r", encoding="utf-8") as f:

  38.                     self.config = json.load(f)

  39.             else:

  40.                 # 使用父类的方法来加载配置

  41.                 self.config = super().load_config()

  42.                 

  43.                 if not self.config:

  44.                     raise Exception("config.json not found")

  45.             # 设置事件处理函数

  46.             self.handlers[Event.ON_HANDLE_CONTEXT] = self.on_handle_context

  47.             self.params_cache = ExpiredDict(300)

  48. 

  49.              # 从配置中提取所需的设置

  50.             self.keys = self.config.get("keys", {})

  51.             self.url_sum = self.config.get("url_sum", {})

  52.             self.search_sum = self.config.get("search_sum", {})

  53.             self.file_sum = self.config.get("file_sum", {})

  54.             self.image_sum = self.config.get("image_sum", {})

  55.             self.note = self.config.get("note", {})

  56. 

  57.             self.sum4all_key = self.keys.get("sum4all_key", "")

  58.             self.search1api_key = self.keys.get("search1api_key", "")

  59.             self.gemini_key = self.keys.get("gemini_key", "")

  60.             self.bibigpt_key = self.keys.get("bibigpt_key", "")

  61.             self.outputLanguage = self.keys.get("outputLanguage", "zh-CN")

  62.             self.opensum_key = self.keys.get("opensum_key", "")

  63.             self.open_ai_api_key = self.keys.get("open_ai_api_key", "")

  64.             self.model = self.keys.get("model", "gpt-3.5-turbo")

  65.             self.open_ai_api_base = self.keys.get("open_ai_api_base", "https://api.openai.com/v1")

  66.             self.xunfei_app_id = self.keys.get("xunfei_app_id", "")

  67.             self.xunfei_api_key = self.keys.get("xunfei_api_key", "")

  68.             self.xunfei_api_secret = self.keys.get("xunfei_api_secret", "")

  69.             self.perplexity_key = self.keys.get("perplexity_key", "")

  70.             self.flomo_key = self.keys.get("flomo_key", "")

  71.             # 之前提示

  72.             self.previous_prompt=''

  73. 

  74.             self.file_sum_enabled = self.file_sum.get("enabled", False)

  75.             self.file_sum_service = self.file_sum.get("service", "")

  76.             self.max_file_size = self.file_sum.get("max_file_size", 15000)

  77.             self.file_sum_group = self.file_sum.get("group", True)

  78.             self.file_sum_qa_prefix = self.file_sum.get("qa_prefix", "问")

  79.             self.file_sum_prompt = self.file_sum.get("prompt", "")

  80. 

  81.             self.image_sum_enabled = self.image_sum.get("enabled", False)

  82.             self.image_sum_service = self.image_sum.get("service", "")

  83.             self.image_sum_group = self.image_sum.get("group", True)

  84.             self.image_sum_qa_prefix = self.image_sum.get("qa_prefix", "问")

  85.             self.image_sum_prompt = self.image_sum.get("prompt", "")

  86. 

  87.             # 初始化成功日志

  88.             logger.info("[file4upload] inited.")

  89.         except Exception as e:

  90.             # 初始化失败日志

  91.             logger.warn(f"file4upload init failed: {e}")

  92. 

  93.     # def on_handle_context(self, e_context: EventContext):

  94.     #     context = e_context["context"]

  95.     #     # logger.info(f'{e_context.__dict__}')

  96.     #     # logger.info('---------------------------------')

  97.     #     # logger.info(f'{ e_context["context"]}')

  98.     #     logger.info('---------------------------------')

  99.     #     logger.info(f'{e_context["context"]["msg"]}')

  100.     #     if context.type not in [ContextType.TEXT, ContextType.SHARING,ContextType.FILE,ContextType.IMAGE]:

  101.     #         return

  102.     #     msg: ChatMessage = e_context["context"]["msg"]

  103.     #     user_id = msg.from_user_id

  104.     #     content = context.content

  105.     #     isgroup = e_context["context"].get("isgroup", False)

  106.     #     print(msg.actual_user_nickname)

  107.     #     itchat.send(f'@{msg.actual_user_nickname}立刻为你服务', toUserName=user_id)

  108. 

  109.     def on_handle_context(self, e_context: EventContext):

  110.         context = e_context["context"]

  111.         # logger.info(f'{e_context.__dict__}')

  112.         # logger.info('---------------------------------')

  113.         # logger.info(f'{ e_context["context"]}')

  114.         # logger.info('---------------------------------')

  115.         # logger.info(f'{e_context["context"]["msg"]}')

  116.         if context.type not in [ContextType.TEXT, ContextType.SHARING,ContextType.FILE,ContextType.IMAGE]:

  117.             return

  118.         msg: ChatMessage = e_context["context"]["msg"]

  119.         user_id = msg.from_user_id

  120.         content = context.content

  121.         isgroup = e_context["context"].get("isgroup", False)

  122.         

  123.         # itchat.send(f'@{msg.actual_user_nickname}立刻为你服务', toUserName=msg.actual_user_nickname)

  124. 

  125.         if isgroup and not self.file_sum_group:

  126.             # 群聊中忽略处理文件

  127.             logger.info("群聊消息，文件处理功能已禁用")

  128.             return

  129.         logger.info("on_handle_context: 处理上下文开始")

  130.         context.get("msg").prepare()

  131. 

  132.         api_key='sk-5z2L4zy9T1w90j6e3T90ANZdyN2zLWClRwFnBzWgzdrG4onx'

  133. 

  134.         

  135.         logger.info(f'当前缓存：self.params_cache:{self.params_cache}') 

  136.         if user_id not in self.params_cache:

  137.             self.params_cache[user_id] = {}

  138.             logger.info(f'初始化缓存:{self.params_cache}') 

  139. 

  140.         if  context.type == ContextType.TEXT and user_id in self.params_cache:

  141.                 self.params_cache[user_id]['previous_prompt']=msg.content

  142. 

  143. 

  144.         # print(f'{msg.__dict__}')

  145.         if context.type == ContextType.IMAGE:

  146.             logger.info('处理图片')

  147.             file_path = context.content

  148.             logger.info(f"on_handle_context: 获取到图片路径 {file_path}，{user_id in self.params_cache}")

  149.             if user_id in self.params_cache:

  150.                 if 'previous_prompt' not in self.params_cache[user_id] and not e_context['context']['isgroup']:

  151. 

  152.                     reply = Reply()

  153.                     reply.type = ReplyType.TEXT

  154.                     reply.content = f"您刚刚上传了一张图片，请问我有什么可以帮您的呢？"

  155.                     e_context["reply"] = reply

  156.                     e_context.action = EventAction.BREAK

  157.                     

  158.                 # else:

  159.                 print(f'准备抽取文字')

  160.                 file_content=extract_content_by_llm(file_path,api_key)

  161. 

  162.                 if file_content is None:

  163.                     logger.info('不能抽取文字,使用图片oss请求LLM')

  164.                     access_key_id = 'LTAI5tRTG6pLhTpKACJYoPR5'

  165.                     access_key_secret = 'E7dMzeeMxq4VQvLg7Tq7uKf3XWpYfN'

  166.                     # OSS区域对应的Endpoint

  167.                     endpoint = 'http://oss-cn-shanghai.aliyuncs.com'  # 根据你的区域选择

  168.                     # Bucket名称

  169.                     bucket_name = 'cow-agent'

  170.                     local_file_path=file_path

  171.                     oss_file_name=f'cow/{os.path.basename(file_path)}'

  172.                     logger.info(f'oss_file_name:{oss_file_name}\n local_file_path :{local_file_path}')

  173.                     file_content = upload_oss(access_key_id, access_key_secret, endpoint, bucket_name, local_file_path, oss_file_name)

  174.                     logger.info(f'写入图片缓存oss 地址{file_content}') 

  175.                 

  176.                 self.params_cache[user_id]['last_content']=file_content                  

  177.     

  178.             # else:

  179.             #     logger.warn(f'还没有建立会话')

  180.             logger.info('删除图片')

  181.             os.remove(file_path)

  182.         

  183.         if context.type == ContextType.FILE:

  184.             logger.info('处理图片')

  185.             file_path = context.content

  186.             logger.info(f"on_handle_context: 获取到文件路径 {file_path}")

  187.             if user_id in self.params_cache:

  188.                 if 'previous_prompt' not in self.params_cache[user_id] and not e_context['context']['isgroup']:

  189.                     reply = Reply()

  190.                     reply.type = ReplyType.TEXT

  191.                     reply.content = f"您刚刚上传了一份文件，请问我有什么可以帮您的呢？"

  192.                     e_context["reply"] = reply

  193.                     e_context.action = EventAction.BREAK

  194.                     

  195.                 # else:

  196.                 print(f'准备抽取文字')

  197.                 file_content=extract_content_by_llm(file_path,api_key)

  198.                 if file_content is None:

  199.                     reply = Reply()

  200.                     reply.type = ReplyType.TEXT

  201.                     reply.content = f"不能处理这份文件"

  202.                     e_context["reply"] = reply

  203.                     e_context.action = EventAction.BREAK

  204.                     return

  205.                 else:

  206.                     self.params_cache[user_id]['last_content']=file_content

  207.             

  208.             logger.info('删除图片')

  209.             os.remove(file_path)

  210. 

  211.         # logger.info('previous_prompt' in self.params_cache[user_id])

  212.         # logger.info('last_content' in self.params_cache[user_id])

  213.         is_previous_prompt='previous_prompt' in self.params_cache[user_id]

  214.         is_last_content='last_content' in self.params_cache[user_id]

  215.         logger.info(f"存在提示词 previous_prompt:{is_previous_prompt}")

  216.         logger.info(f'存在内容 last_content:{is_last_content}' )

  217.         

  218. 

  219.         if 'previous_prompt' in self.params_cache[user_id] and 'last_content' in self.params_cache[user_id] and contains_keywords(self.params_cache[user_id]['previous_prompt']):

  220.             #先回应

  221.             logger.info('先回应')

  222.             # reply2 = Reply()

  223.             # reply2.type = ReplyType.TEXT

  224.             # reply2.content = f"已经收到，立刻为你服务"

  225.             # msg:ChatMessage = e_context['context']['msg']

  226.             # e_context['reply'] = reply2

  227.             # e_context.action = EventAction.BREAK # 事件结束

  228. 

  229. 

  230.             # reply = Reply()

  231.             # reply.type = ReplyType.TEXT

  232.             # reply.content = f"已经收到，立刻为你服务"

  233.             # e_context["reply"] = reply

  234.             # e_context.action = EventAction.BREAK

  235.             receiver=user_id

  236.             print(receiver)

  237.             # itchat_content=  '' if e_context['context']['isgroup'] else '[小蕴]'+"已经收到，立刻为你服务"

  238.             # if e_context['context']['isgroup']:

  239.             #     itchat_content =f'@{msg.actual_user_nickname}已经收到，立刻为你服务'

  240.             # else:

  241.             #     itchat_content = '[小蕴]'+"已经收到，立刻为你服务"

  242. 

  243.             text=self.params_cache[user_id]['previous_prompt']

  244.             logger.info(f'{text},{contains_keywords(text)}')

  245. 

  246. 

  247.             itchat_content=  f'@{msg.actual_user_nickname}' if e_context['context']['isgroup'] else '[小蕴]'

  248.             itchat_content+="已经收到，立刻为您服务"

  249.             flag=contains_keywords(text)

  250.             if flag==True:

  251.                 print('发送'+itchat_content)

  252.                 itchat.send(itchat_content, toUserName=receiver)

  253. 

  254.             e_context.action = EventAction.BREAK

  255. 

  256.         if 'previous_prompt' in self.params_cache[user_id] and 'last_content' in self.params_cache[user_id]:

  257. 

  258.             if  contains_keywords(self.params_cache[user_id]['previous_prompt']):

  259.                 e_context["context"].type = ContextType.TEXT

  260.                 e_context["context"].content ="<content>"+self.params_cache[user_id]['last_content']+"</content>"+'\n\t'+"<ask>"+self.params_cache[user_id]['previous_prompt']+"</ask>"

  261.                 logger.info(f'conze4upload 插件处理上传文件或图片')

  262.                 e_context.action = EventAction.CONTINUE

  263.                 

  264.                 # 清空清空缓存

  265.                 self.params_cache.clear()

  266.                 logger.info(f'清空缓存后：{self.params_cache}')

  267.             else:

  268.                 if not e_context['context']['isgroup']:

  269.                     reply = Reply()

  270.                     reply.type = ReplyType.TEXT

  271.                     # reply.content = f"{remove_markdown(reply_content)}\n\n💬5min内输入{self.file_sum_qa_prefix}+问题，可继续追问" 

  272.                     reply.content = f"您刚刚上传了，请问我有什么可以帮您的呢？"

  273.                     e_context["reply"] = reply

  274.                 e_context.action = EventAction.BREAK

  275.                 return

  276. 

  277. 

  278.         ## e_context.action = EventAction.BREAK  

  279.        

  280. 

  281. def remove_markdown(text):

  282.     # 替换Markdown的粗体标记

  283.     text = text.replace("**", "")

  284.     # 替换Markdown的标题标记

  285.     text = text.replace("### ", "").replace("## ", "").replace("# ", "")

  286.     return text

  287. 

  288. def extract_content_by_llm(file_path: str, api_key: str) -> str:

  289.     logger.info(f'大模型开始抽取文字')

  290.     try:

  291.         headers = {

  292.             'Authorization': f'Bearer {api_key}'

  293.         }

  294.         data = {

  295.             'purpose': 'file-extract',

  296.         }

  297.         file_name=os.path.basename(file_path)

  298.         files = {

  299.             'file': (file_name, open(Path(file_path), 'rb')),

  300.         }

  301.         # print(files)

  302.         api_url='https://api.moonshot.cn/v1/files'

  303.         response = requests.post(api_url, headers=headers, files=files, data=data)

  304.         response_data = response.json()

  305.         file_id = response_data.get('id')

  306.         response=requests.get(url=f"https://api.moonshot.cn/v1/files/{file_id}/content", headers=headers)

  307.         print(response.text)

  308.         response_data = response.json()

  309.         content = response_data.get('content')

  310.         return content

  311.     except requests.exceptions.RequestException as e:

  312.         logger.error(f"Error calling LLM API: {e}")

  313.         return None

  314. 

  315. def upload_oss(access_key_id, access_key_secret, endpoint, bucket_name, local_file_path, oss_file_name, expiration_days=7):

  316.     """

  317.     上传文件到阿里云OSS并设置生命周期规则，同时返回文件的公共访问地址。

  318.     

  319.     :param access_key_id: 阿里云AccessKey ID

  320.     :param access_key_secret: 阿里云AccessKey Secret

  321.     :param endpoint: OSS区域对应的Endpoint

  322.     :param bucket_name: OSS中的Bucket名称

  323.     :param local_file_path: 本地文件路径

  324.     :param oss_file_name: OSS中的文件存储路径

  325.     :param expiration_days: 文件保存天数，默认7天后删除

  326.     :return: 文件的公共访问地址

  327.     """

  328.     

  329.     # 创建Bucket实例

  330.     auth = oss2.Auth(access_key_id, access_key_secret)

  331.     bucket = oss2.Bucket(auth, endpoint, bucket_name)

  332. 

  333.     ### 1. 设置生命周期规则 ###

  334.     rule_id = f'delete_after_{expiration_days}_days'  # 规则ID

  335.     prefix = oss_file_name.split('/')[0] + '/'  # 设置规则应用的前缀为文件所在目录

  336. 

  337.     # 定义生命周期规则

  338.     rule = oss2.models.LifecycleRule(rule_id, prefix, status=oss2.models.LifecycleRule.ENABLED,

  339.                                      expiration=oss2.models.LifecycleExpiration(days=expiration_days))

  340. 

  341.     # 设置Bucket的生命周期

  342.     lifecycle = oss2.models.BucketLifecycle([rule])

  343.     bucket.put_bucket_lifecycle(lifecycle)

  344. 

  345.     print(f"已设置生命周期规则：文件将在{expiration_days}天后自动删除")

  346. 

  347.     ### 2. 上传文件到OSS ###

  348.     bucket.put_object_from_file(oss_file_name, local_file_path)

  349. 

  350.     ### 3. 构建公共访问URL ###

  351.     file_url = f"http://{bucket_name}.{endpoint.replace('http://', '')}/{oss_file_name}"

  352. 

  353.     print(f"文件上传成功，公共访问地址：{file_url}")

  354.     

  355.     return file_url

  356.  

  357. def contains_keywords_by_re(text):

  358.     # 匹配<ask>标签中的内容

  359.     # match = re.search(r'<ask>(.*?)</ask>', text)

  360.     match = re.search(r'(.*?)', text)

  361.     if match:

  362.         content = match.group(1)

  363.         # 检查关键词

  364.         keywords = ['分析', '总结', '报告', '描述']

  365.         for keyword in keywords:

  366.             if keyword in content:

  367.                 return True

  368.     return False

  369. 

  370. def contains_keywords(text):

  371.     keywords = ["分析", "总结", "报告", "描述"]

  372.     return any(keyword in text for keyword in keywords)