Skip to main content

lark(feishu) websocket client

Project description

Feishu-Webhook-Proxy

  1. 将飞书webhook代理成websocket
  2. 企业自建应用不用创建公网的回调地址,直接本地使用websocket客户端连上这个转发地址

设计

  1. 使用nchan维护websocket的连接
  2. 将飞书的回调消息,抽取飞书相关的头信息,外面包一层json,使用X-Request-Id作为唯一ID,推送给对应的channel,如果连接对应websocket的客户端回复了X-Request-Id对应的消息,就回复给飞书(这里主要用于第一次配置回调)
  3. 客户端自己保存飞书的密钥信息,从转发服务走的消息都是加密的。
  4. 客户端调用飞书其他接口,直接走自己的网络

安全性

  1. 飞书回调消息都是加密的,只能由websocket客户端自己解密,转发服务是透明的。
  2. 如何确保自己的channel不会被别人恶意使用?

使用nginx basic auth,nchan支持auth_request,在对应的request里面使用basic auth就能做校验

实现

  • 部署一个nchan(openresty版本)
  • 配置一个internal的location,给内部转发飞书消息使用
  • 配置一个location,作为飞书webhook转发(处理消息转发逻辑,如果是配置连接,就重定向到request_id对应的channel等待客户端返回challenge给飞书)

organization

使用一个organization对当前组织下面的所有bot进行管理 这样所有的消息可以通过org_<name>一个channel推送,这种模式下启动服务的时候可以不用提前注册所有的bot,可以动态的加入新的bot进去

  • 对org_的channel增加basic auth
  • hook链接转发消息兼容organization
  • 新增一个支持organization的client
client = Client(bot1, bot2, org_name='org_lloyd', org_passwd='passwd')

使用

python sdk

pip install wslarkbot

from wslarkbot import *

class MyBot(Bot):
    def on_message(self, data, raw_message, **kwargs):
        # 定义每一个机器人拿到消息后的处理逻辑
        print('on_message', self.app_id, data, raw_message)
        if 'header' in data:
            if data['header']['event_type'] == 'im.message.receive_v1' and data['event']['message']['message_type'] == 'text':
                message_id = data['event']['message']['message_id']
                content = json.loads(data['event']['message']['content'])
                text = content['text']
                # 测试回复消息,初始化bot的时候,需要配置app_secret才能发出去消息
                self.reply_text(message_id, 'reply: ' + text)
                # 回复卡片消息
                self.reply_card(message_id, FeishuMessageCard(
                    FeishuMessageDiv('reply'),
                    FeishuMessageHr(),
                    FeishuMessageDiv(text),
                    FeishuMessageNote(FeishuMessagePlainText('🤖'))
                ))

bot1 = MyBot('cli_xxx', app_secret='xxx', encrypt_key='xxx')
bot2 = MyBot('cli_xxx', app_secret='xxx', encrypt_key='xxx')

# 一个websocket连接,支持同时监听多个机器人回调消息
client = Client(bot1, bot2)
client.start()

集成openai

test_openai.py文件中

  1. 继承Bot增加自己处理消息的回调
class TextMessageBot(Bot):
    def on_message(self, data, *args, **kwargs):
        if 'header' in data:
            if data['header']['event_type'] == 'im.message.receive_v1' and data['event']['message']['message_type'] == 'text':
                content = json.loads(data['event']['message']['content'])
                if self.app:
                    return self.app.process_text_message(text=content['text'], **data['event']['message'])


  1. 写一个应用:处理文本消息
class Application(object):
    def process_text_message(self, text, message_id, **kwargs):
        if text == '/help' or text == '帮助':
            self.bot.reply_card(
                message_id,
                FeishuMessageCard(
                    FeishuMessageDiv('👋 你好呀,我是一款基于OpenAI技术的智能聊天机器人'),
                    FeishuMessageHr(),
                    FeishuMessageDiv('🎒 **需要更多帮助**\n文本回复 *帮助* 或 */help*', tag='lark_md'),
                    header=FeishuMessageCardHeader('🎒需要帮助吗?'),
                )
            )
        elif text:
            chat = ChatOpenAI(
                callbacks=[OpenAICallbackHandler(self.bot, message_id)],
                **self.openai_options
            )
            system_message = [SystemMessage(content=self.system_role)] if self.system_role else []
            chat_history = []  # TODO
            messages = system_message + chat_history + [HumanMessage(content=text)]
            message = chat(messages)
            logging.debug("reply message %r", message)
        else:
            logging.warn("empty text", text)
  1. 初始化应用,启动机器人
if __name__ == "__main__":
    app = Application(
        openai_api_base='',
        openai_api_key='',
        app_id='',
        app_secret='',
        encrypt_key='',
        verification_token='',
    )
    client = Client(app.bot)
    client.start(True)  # debug mode

运行示例

pip install wslarkbot langchain openai click
python test_openai.py

image

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

wslarkbot-0.2.3.tar.gz (10.2 kB view details)

Uploaded Source

File details

Details for the file wslarkbot-0.2.3.tar.gz.

File metadata

  • Download URL: wslarkbot-0.2.3.tar.gz
  • Upload date:
  • Size: 10.2 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/4.0.2 CPython/3.11.6

File hashes

Hashes for wslarkbot-0.2.3.tar.gz
Algorithm Hash digest
SHA256 1e9a88a0bb96d5bb491c043c865e54c96469ef223c71d3369243b66c96e6218b
MD5 7f966704c8fdf482d4ca033917243dbc
BLAKE2b-256 267f5b0cd2070cb52e33b3ff562d5c8f2583bda16b8326f503daeda34f2052b9

See more details on using hashes here.

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page