mirror of
https://github.com/Zippland/Bubbles.git
synced 2026-03-04 07:47:48 +08:00
269 lines
11 KiB
Python
269 lines
11 KiB
Python
#! /usr/bin/env python3
|
||
# -*- coding: utf-8 -*-
|
||
|
||
import logging
|
||
import base64
|
||
import os
|
||
from datetime import datetime
|
||
|
||
import httpx
|
||
from openai import APIConnectionError, APIError, AuthenticationError, OpenAI
|
||
|
||
|
||
class ChatGPT():
|
||
def __init__(self, conf: dict) -> None:
|
||
key = conf.get("key")
|
||
api = conf.get("api")
|
||
proxy = conf.get("proxy")
|
||
prompt = conf.get("prompt")
|
||
self.model = conf.get("model", "gpt-3.5-turbo")
|
||
self.LOG = logging.getLogger("ChatGPT")
|
||
if proxy:
|
||
self.client = OpenAI(api_key=key, base_url=api, http_client=httpx.Client(proxy=proxy))
|
||
else:
|
||
self.client = OpenAI(api_key=key, base_url=api)
|
||
self.conversation_list = {}
|
||
self.system_content_msg = {"role": "system", "content": prompt}
|
||
# 确认是否使用支持视觉的模型
|
||
self.support_vision = self.model == "gpt-4-vision-preview" or self.model == "gpt-4o" or "-vision" in self.model
|
||
|
||
def __repr__(self):
|
||
return 'ChatGPT'
|
||
|
||
@staticmethod
|
||
def value_check(conf: dict) -> bool:
|
||
if conf:
|
||
if conf.get("key") and conf.get("api") and conf.get("prompt"):
|
||
return True
|
||
return False
|
||
|
||
def get_answer(self, question: str, wxid: str, system_prompt_override=None) -> str:
|
||
# wxid或者roomid,个人时为微信id，群消息时为群id
|
||
|
||
# 检查是否是新对话
|
||
is_new_conversation = wxid not in self.conversation_list
|
||
|
||
# 保存临时系统提示的状态
|
||
temp_system_used = False
|
||
original_prompt = None
|
||
|
||
if system_prompt_override:
|
||
# 只有新对话才临时修改系统提示
|
||
if is_new_conversation:
|
||
# 临时保存原始系统提示，以便可以恢复
|
||
original_prompt = self.system_content_msg["content"]
|
||
# 设置临时系统提示
|
||
self.system_content_msg["content"] = system_prompt_override
|
||
temp_system_used = True
|
||
self.LOG.debug(f"为新对话 {wxid} 临时设置系统提示")
|
||
else:
|
||
# 对于已存在的对话，我们将在API调用时临时使用覆盖提示，而不修改对话历史
|
||
self.LOG.debug(f"对话 {wxid} 已存在，系统提示覆盖将仅用于本次API调用")
|
||
|
||
# 添加用户问题到对话历史
|
||
self.updateMessage(wxid, question, "user")
|
||
|
||
# 如果修改了系统提示，现在恢复它
|
||
if temp_system_used and original_prompt is not None:
|
||
self.system_content_msg["content"] = original_prompt
|
||
self.LOG.debug(f"已恢复默认系统提示")
|
||
|
||
rsp = ""
|
||
try:
|
||
# 准备API调用的消息列表
|
||
api_messages = []
|
||
|
||
# 对于已存在的对话，临时应用系统提示覆盖（如果有）
|
||
if not is_new_conversation and system_prompt_override:
|
||
# 第一个消息可能是系统提示
|
||
has_system = self.conversation_list[wxid][0]["role"] == "system"
|
||
|
||
# 使用临时系统提示替代原始系统提示
|
||
if has_system:
|
||
# 复制除了系统提示外的所有消息
|
||
api_messages = [{"role": "system", "content": system_prompt_override}]
|
||
api_messages.extend(self.conversation_list[wxid][1:])
|
||
else:
|
||
# 如果没有系统提示，添加一个
|
||
api_messages = [{"role": "system", "content": system_prompt_override}]
|
||
api_messages.extend(self.conversation_list[wxid])
|
||
else:
|
||
# 对于新对话或没有临时系统提示的情况，使用原始对话历史
|
||
api_messages = self.conversation_list[wxid]
|
||
|
||
# o系列模型不支持自定义temperature，只能使用默认值1
|
||
params = {
|
||
"model": self.model,
|
||
"messages": api_messages
|
||
}
|
||
|
||
# 只有非o系列模型才设置temperature
|
||
if not self.model.startswith("o"):
|
||
params["temperature"] = 0.2
|
||
|
||
ret = self.client.chat.completions.create(**params)
|
||
rsp = ret.choices[0].message.content
|
||
rsp = rsp[2:] if rsp.startswith("\n\n") else rsp
|
||
rsp = rsp.replace("\n\n", "\n")
|
||
self.updateMessage(wxid, rsp, "assistant")
|
||
except AuthenticationError:
|
||
self.LOG.error("OpenAI API 认证失败，请检查 API 密钥是否正确")
|
||
except APIConnectionError:
|
||
self.LOG.error("无法连接到 OpenAI API，请检查网络连接")
|
||
except APIError as e1:
|
||
self.LOG.error(f"OpenAI API 返回了错误：{str(e1)}")
|
||
rsp = "无法从 ChatGPT 获得答案"
|
||
except Exception as e0:
|
||
self.LOG.error(f"发生未知错误：{str(e0)}")
|
||
rsp = "无法从 ChatGPT 获得答案"
|
||
|
||
return rsp
|
||
|
||
def encode_image_to_base64(self, image_path: str) -> str:
|
||
"""将图片文件转换为Base64编码
|
||
|
||
Args:
|
||
image_path (str): 图片文件路径
|
||
|
||
Returns:
|
||
str: Base64编码的图片数据
|
||
"""
|
||
try:
|
||
with open(image_path, "rb") as image_file:
|
||
return base64.b64encode(image_file.read()).decode('utf-8')
|
||
except Exception as e:
|
||
self.LOG.error(f"图片编码失败: {str(e)}")
|
||
return ""
|
||
|
||
def get_image_description(self, image_path: str, prompt: str = "请详细描述这张图片中的内容") -> str:
|
||
"""使用GPT-4 Vision分析图片内容
|
||
|
||
Args:
|
||
image_path (str): 图片文件路径
|
||
prompt (str, optional): 提示词. 默认为"请详细描述这张图片中的内容"
|
||
|
||
Returns:
|
||
str: 模型对图片的描述
|
||
"""
|
||
if not self.support_vision:
|
||
self.LOG.error(f"当前模型 {self.model} 不支持图片理解，请使用gpt-4-vision-preview或gpt-4o")
|
||
return "当前模型不支持图片理解功能，请联系管理员配置支持视觉的模型（如gpt-4-vision-preview或gpt-4o）"
|
||
|
||
if not os.path.exists(image_path):
|
||
self.LOG.error(f"图片文件不存在: {image_path}")
|
||
return "无法读取图片文件"
|
||
|
||
try:
|
||
base64_image = self.encode_image_to_base64(image_path)
|
||
if not base64_image:
|
||
return "图片编码失败"
|
||
|
||
# 构建带有图片的消息
|
||
messages = [
|
||
{"role": "system", "content": "你是一个图片分析专家，擅长分析图片内容并提供详细描述。"},
|
||
{
|
||
"role": "user",
|
||
"content": [
|
||
{"type": "text", "text": prompt},
|
||
{
|
||
"type": "image_url",
|
||
"image_url": {
|
||
"url": f"data:image/jpeg;base64,{base64_image}"
|
||
}
|
||
}
|
||
]
|
||
}
|
||
]
|
||
|
||
# 使用GPT-4 Vision模型
|
||
params = {
|
||
"model": self.model,
|
||
"messages": messages,
|
||
"max_tokens": 1000 # 限制输出长度
|
||
}
|
||
|
||
# 支持视觉的模型可能有不同参数要求
|
||
if not self.model.startswith("o"):
|
||
params["temperature"] = 0.7
|
||
|
||
response = self.client.chat.completions.create(**params)
|
||
description = response.choices[0].message.content
|
||
description = description[2:] if description.startswith("\n\n") else description
|
||
description = description.replace("\n\n", "\n")
|
||
|
||
return description
|
||
|
||
except AuthenticationError:
|
||
self.LOG.error("OpenAI API 认证失败，请检查 API 密钥是否正确")
|
||
return "API认证失败，无法分析图片"
|
||
except APIConnectionError:
|
||
self.LOG.error("无法连接到 OpenAI API，请检查网络连接")
|
||
return "网络连接错误，无法分析图片"
|
||
except APIError as e1:
|
||
self.LOG.error(f"OpenAI API 返回了错误：{str(e1)}")
|
||
return f"API错误：{str(e1)}"
|
||
except Exception as e0:
|
||
self.LOG.error(f"分析图片时发生未知错误：{str(e0)}")
|
||
return f"处理图片时出错：{str(e0)}"
|
||
|
||
def updateMessage(self, wxid: str, content: str, role: str) -> None:
|
||
now_time = str(datetime.now().strftime("%Y-%m-%d %H:%M:%S"))
|
||
|
||
time_mk = "当需要回答时间时请直接参考回复:"
|
||
# 初始化聊天记录,组装系统信息
|
||
if wxid not in self.conversation_list.keys():
|
||
# 此时self.system_content_msg可能已经被get_answer临时修改
|
||
# 但这没关系，因为在get_answer结束前会恢复
|
||
question_ = [
|
||
self.system_content_msg,
|
||
{"role": "system", "content": "" + time_mk + now_time}
|
||
]
|
||
self.conversation_list[wxid] = question_
|
||
|
||
# 当前问题或回答
|
||
content_message = {"role": role, "content": content}
|
||
self.conversation_list[wxid].append(content_message)
|
||
|
||
# 更新时间标记
|
||
for cont in self.conversation_list[wxid]:
|
||
if cont["role"] != "system":
|
||
continue
|
||
if cont["content"].startswith(time_mk):
|
||
cont["content"] = time_mk + now_time
|
||
|
||
# 控制对话历史长度
|
||
# 只存储10条记录，超过滚动清除
|
||
max_history = 12 # 包括1个系统提示和1个时间标记
|
||
i = len(self.conversation_list[wxid])
|
||
if i > max_history:
|
||
# 计算需要删除多少条记录
|
||
if self.conversation_list[wxid][0]["role"] == "system" and self.conversation_list[wxid][1]["role"] == "system":
|
||
# 如果前两条都是系统消息，保留它们，删除较早的用户和助手消息
|
||
to_delete = i - max_history
|
||
del self.conversation_list[wxid][2:2+to_delete]
|
||
self.LOG.debug(f"滚动清除微信记录：{wxid}，删除了{to_delete}条历史消息")
|
||
else:
|
||
# 如果结构不符合预期，简单地保留最近的消息
|
||
self.conversation_list[wxid] = self.conversation_list[wxid][-max_history:]
|
||
self.LOG.debug(f"滚动清除微信记录：{wxid}，只保留最近{max_history}条消息")
|
||
|
||
|
||
if __name__ == "__main__":
|
||
from configuration import Config
|
||
config = Config().CHATGPT
|
||
if not config:
|
||
exit(0)
|
||
|
||
chat = ChatGPT(config)
|
||
|
||
while True:
|
||
q = input(">>> ")
|
||
try:
|
||
time_start = datetime.now() # 记录开始时间
|
||
print(chat.get_answer(q, "wxid"))
|
||
time_end = datetime.now() # 记录结束时间
|
||
|
||
print(f"{round((time_end - time_start).total_seconds(), 2)}s") # 计算的时间差为程序的执行时间，单位为秒/s
|
||
except Exception as e:
|
||
print(e)
|