unfuck VS thefuck

This is sweet! This is actually a very similar approach to how I deobfuscate Python bytecode: https://github.com/landaire/unfuck/blob/bfa164b4e261deffeb37...

My code is pretty messy, but I take the same exact approach of taking known function parameters, interpreting the instructions, and removing any condition and the instructions which built its arguments if it evaluates to a constant value. Even called it partial execution as well :p

I haven't heard of this tool before, but I don't think it would work for obfuscated code. Check out the graph image on my wiki [1]. While this image doesn't show the exact scenario, imagine that the first instruction is a `JUMP_ABSOLUTE 100` and the rest of the instructions between offset [3,100) are just garbage or invalid.

A naive disassembler (like the `dis` module in python) interprets the bytecode linearly -- i.e. one instruction after another. Rizin's diassembler [2] seems to take the same approach. The way I do disassembly is to only disassemble code paths that are potentially executed by queueing non-conditional jumps, both targets of a conditional jump, and the next instruction when the current instruction is non-jumping instruction.

[1] https://github.com/landaire/unfuck/wiki/Obfuscation-Tricks

Are you by chance interested in a command named after the four-letter word, which automatically fixes and reruns the last command: https://github.com/nvbn/thefuck

With a good terminal, you can work 2-4 times faster. This will save time and reduced your errors. You can also use fuck (just like how you swear most of the time) to correct errors easily.

Now I want a thefuck implementation via uncensored LLMs.

>spellcheck on commands

I prefer to just type "fuck":

https://github.com/nvbn/thefuck

I am starting to learn Rust and I am starting to implement the fuck CLI tool in Rust. Do you think this is a good use of my learning time?