haibersut/claude-web

一个基于 Python 用于和 Claude Web API 进行交互的程序

claude-web's Issues

TODO: 修复文件转文字请求错误

请求内容：

POST https://claude.ai/api/convert_document HTTP/1.1
'Host': 'claude.ai'
Connection: keep-alive
Content-Length: xxxxxxxxx
sec-ch-ua: xxxxxx
sec-ch-ua-platform: "Windows"
sec-ch-ua-mobile: ?0
User-Agent: xxxxxxx
'Content-Type: multipart/form-data'; boundary=----xxxxx
'Accept': '/',
'Origin': 'https://claude.ai',
Sec-Fetch-Site: same-origin
Sec-Fetch-Mode: cors
Sec-Fetch-Dest: empty
'Referer': 'https://claude.ai/chats',
'Accept-Encoding': 'gzip, deflate, br',
Accept-Language: xxxxxxx
Cookie: sessionKey=xxxx

------xxxxxx
Content-Disposition: form-data; name="file"; filename="test.pdf"
Content-Type: application/pdf

（下面是PDF的内容）

---

响应内容：

HTTP/1.1 200 OK
Date: xxxxxxx
Content-Type: application/json
Connection: keep-alive
strict-transport-security: max-age=xxxx; includeSubDomains; preload
x-xss-protection: 1; mode=block
x-frame-options: xxxx
x-content-type-options: nosniff
content-security-policy: xxxxxxx
vary: RSC, Next-Router-State-Tree, Next-Router-Prefetch, Accept-Encoding
Cache-Control: private
Via: 1.1 google
CF-Cache-Status: DYNAMIC
Set-Cookie: xxxxxxxxx
Server: cloudflare
CF-RAY: xxxxxx
Content-Length: xxxxxx

{"file_name":"test.pdf","file_size":xxxxx,"file_type":"pdf","extracted_content":"PDF文件中的文字内容.","totalPages":xxxxx}

其实可以自行解析构造即可，返回的内容会被加入到"attachments": []中，然后跟随文本一起请求。