update README.md

This commit is contained in:
LanQian 2024-04-23 03:09:45 +08:00
parent cf2cff84de
commit ed2495582d
2 changed files with 29 additions and 34 deletions

View File

@ -24,7 +24,7 @@
> - [x] ArkoseToken
> - [x] GPT4.0 画图、工具 (测试中可能有bug)
> - [x] 使用 RefreshToken 代替 AccessToken
> - [x] 反向代理 UI (http://127.0.0.1:5005, 不支持登录)
> - [x] 反向代理 UI (http://127.0.0.1:5005, 不支持登录使用)
>
> TODO
>
@ -42,7 +42,7 @@
git clone https://github.com/LanQian528/chat2api
cd chat2api
pip install -r requirements.txt
python chat2api.py
python app.py
```
### Docker部署
@ -139,13 +139,12 @@ curl --location 'http://127.0.0.1:5005/v1/chat/completions' \
### 环境变量
```
AUTHORIZATION=your_first_token, your_second_token // 使用免登3.5的Bearer token不设置则无需Bearer token
FREE35_BASE_URL=https://chat.openai.com/backend-anon // 免登3.5的网关地址
CHATGPT_BASE_URL=https://chat.openai.com/backend-api // 非免登3.5、4.0的网关地址
HISTORY_DISABLED=true // 是否保存聊天记录
PROXY_URL=your_first_proxy, your_second_proxy // 代理url多个代理用逗号分隔
ARKOSE_TOKEN_URL=https://arkose.example.com/token // 获取Arkose token的地址上文有提供说明
RETRY_TIMES=3 // 重试次数
AUTHORIZATION=your_first_token, your_second_token // 使用免登3.5的Bearer token不设置则无需Bearer token
CHATGPT_BASE_URL=https://chat.openai.com // ChatGPT网关地址
HISTORY_DISABLED=true // 是否保存聊天记录
PROXY_URL=your_first_proxy, your_second_proxy // 代理url多个代理用逗号分隔
ARKOSE_TOKEN_URL=https://arkose.example.com/token // 获取Arkose token的地址上文有提供说明
RETRY_TIMES=3 // 重试次数
```
## License

View File

@ -5,7 +5,6 @@ from fastapi.responses import StreamingResponse, Response
from utils.Client import Client
from utils.config import chatgpt_base_url_list, proxy_url_list
headers_reject_list = [
"x-real-ip",
"x-forwarded-for",
@ -57,9 +56,9 @@ headers_reject_list = [
"cf-visitor",
]
async def chatgpt_reverse_proxy(request: Request, path: str):
try:
base_url = random.choice(chatgpt_base_url_list)
origin_host = request.url.netloc
if ":" in origin_host:
petrol = "http"
@ -67,30 +66,30 @@ async def chatgpt_reverse_proxy(request: Request, path: str):
petrol = "https"
if path.startswith("v1/"):
base_url = "https://ab.chatgpt.com"
if path.startswith("authorize") and "max_age" not in dict(request.query_params).keys():
base_url = "https://auth.openai.com"
if path.startswith("authorize") or path.startswith("u/") or path.startswith("oauth/") or path.startswith("assets/"):
base_url = "https://auth0.openai.com"
else:
base_url = random.choice(chatgpt_base_url_list)
params = dict(request.query_params)
headers = {
key: value for key, value in request.headers.items()
if (key.lower() not in ["host", "origin", "referer", "user-agent", "authorization"] and key.lower() not in headers_reject_list)
if (key.lower() not in ["host", "origin", "referer", "user-agent",
"authorization"] and key.lower() not in headers_reject_list)
}
cookies = dict(request.cookies)
headers.update({
"Accept-Language": "en-US,en;q=0.9",
"Host": f"{base_url.split('//')[1]}",
"Origin": f"{base_url}",
"Referer": f"{base_url}/{path}",
"Sec-Ch-Ua": '"Chromium";v="123", "Not(A:Brand";v="24", "Microsoft Edge";v="123"',
"Sec-Ch-Ua-Mobile": "?0",
"Sec-Ch-Ua-Platform": "\"Windows\"",
"Sec-Fetch-Dest": "empty",
"Sec-Fetch-Mode": "cors",
"Sec-Fetch-Site": "same-origin",
"User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/123.0.0.0 Safari/537.36 Edg/123.0.0.0"
"accept-Language": "en-US,en;q=0.9",
"host": base_url.replace("https://", "").replace("http://", ""),
"origin": base_url,
"referer": f"{base_url}/",
"sec-ch-ua": '"Chromium";v="124", "Microsoft Edge";v="124", "Not-A.Brand";v="99"',
"sec-ch-ua-mobile": "?0",
"sec-ch-ua-platform": "\"Windows\"",
"sec-fetch-dest": "empty",
"sec-fetch-mode": "cors",
"sec-fetch-site": "same-origin",
"user-agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/124.0.0.0 Safari/537.36 Edg/124.0.0.0"
})
if request.headers.get('Authorization'):
headers['Authorization'] = request.headers['Authorization']
@ -103,23 +102,20 @@ async def chatgpt_reverse_proxy(request: Request, path: str):
r = await client.request(request.method, f"{base_url}/{path}", params=params, headers=headers, cookies=cookies,
data=data, stream=True)
if 'stream' in r.headers.get("content-type", ""):
if r.status_code == 302:
return Response(status_code=302, headers={"Location": r.headers.get("Location")})
elif 'stream' in r.headers.get("content-type", ""):
return StreamingResponse(r.aiter_content(), media_type=r.headers.get("content-type", ""))
else:
content = ((await r.atext()).replace("chat.openai.com", origin_host)
.replace("ab.chatgpt.com", origin_host)
.replace("cdn.oaistatic.com", origin_host)
.replace("auth.openai.com", origin_host)
.replace("auth0.openai.com", origin_host)
.replace("https", petrol))
response = Response(content=content, media_type=r.headers.get("content-type"), status_code=r.status_code)
for key, value in r.cookies.items():
if key in cookies.keys():
continue
if key.startswith("__"):
response.set_cookie(key=key, value=value, secure=True, httponly=True, path='/')
else:
response.set_cookie(key=key, value=value, secure=True, httponly=True)
response.set_cookie(key=key, value=value)
return response
except Exception as e: