Structured output

[wasamasa]

I'm using remote-method-guesser to scan several hosts and dissecting the textual output is tedious. So I've looked at the code and only got a vague idea (my Java skills are pretty bad) how to allow for alternative output formats:

• Restructure code so that there's a dedicated logger interface with two implementations, a TTY one using ANSI escapes and a JSON one
• Restructure code so that the logger interface methods are used, which receive a context (current host/port/probe) and event (with all the details logged to the console)

Seems like quite a bit of work, but on the upside, it would make the existing TTY formatting less terrible.

The JSON output would look something like this

{"host": "...", "port": "...", "probe": "rmi_registry_jep290_check", "status": "not vulnerable"}
{"host": "...", "port": "...", "probe": "rmi_registry_jep290_bypass", "status": "vulnerable", "gadget": "An Trinh"}

[qtc-de]

Hi @wasamasa

this is definitely true and was already requested in #38 (that was somehow closed). rmg was never designed to run at scale and has some poor design decisions when it comes to running against multiple hosts. As you say, it would require quite some work to do this correctly and my time is currently quite limited.

However, hacking something together that provides JSON output when running against one host should be quite simple. I will work on it 😉