Inspired by: Brainfuck/brainfuck-visualizer-master
There is the smallest interpreter interpreters/smallest.js.
And the interpreters/complete.js with more features for web use cases.
- Yields the result of each char before interpreting it
- Receives a object with the last result to start from
const bf = bfinterpreter('++++++++[>++++[>++>+++>+++>+<<<<-]>+>+>->>+[<]<-]>>.>---.+++++++..+++.>>.<-.<.+++.------.--------.>>+.>++.!', { pause: true })
while (!(a = bf.next()).done)
console.log(a.value)It translates text to Brainfuck.
depreterv2.js
Same configuration as #interpreter 🐍
A really smart depreter that writes code with the same techniques I myself use when I manually writing Brainf#ck.
It chooses the smallest result from many techniques to get to the number on the memory with less code as possible.
It uses:
- Search for the cell with the closest number to the number I want, so we can have less instructions than we would need starting from zero.
><[-]+++++++and><[+]-------
This technique is used when the smallest way to get to the number is starting from zero.- Try creating a new cell to see if the code gets smaller.
- Loops
- In loops we lock the chosen cell (1st technique) so that the counter code doesn't choose it.
- We choose the smallest factors of the number
Note
TODOS:
- Support loops inside loops
- Support loop counter to contain a loop
Its output can be redirected to a .s file, compiled with GNU Assembly,
and then ran sending the output to xxd to show the result in hex.
node main.js > prog.s
./comp.sh prog && ./prog | xxd❗ Different from #interpreter this works with ASCII (0-127), not UTF-16 (fromCharCode). Each memory cell is 1 byte!
compilers/jit
Inspired by: tsoding/bfjit
It compiles Brainfuck to a valid ELF64 format entirely in JavaScript, and runs it with a C shared library.
Just like #as-transpiler, it works with ASCII, so be careful.
Soon.