Proxy format converter

Formats it reads and writes

It works out which part is the host and which is the password by looking for a valid IP or hostname next to a port number, so a list that mixes layouts converts fine.

• ip:port:user:pass, the layout most proxy sellers hand out
• user:pass@ip:port, what most scrapers and libraries expect after the scheme
• user:pass:ip:port and ip:port@user:pass, used by some bots and checkers
• http://user:pass@ip:port and socks5://user:pass@ip:port, full URLs for code and browser extensions
• ip:port, credentials stripped, for tools that ask for them in separate fields

Which format does my tool want?

• Antidetect browsers like AdsPower, GoLogin and Multilogin: most take ip:port:user:pass directly, or have separate fields.
• curl: -x http://user:pass@ip:port
• Python requests: a proxies dict whose value is http://user:pass@ip:port
• Puppeteer and Playwright: the server as http://ip:port, with username and password passed separately.

Passwords with characters like @ or : are percent-encoded in the URL formats, since they'd break the URL otherwise. The plain layouts leave them as they are.

Once they're converted, run them through the proxy checker to see which ones work and what IP type they exit from.