In this chapter, we first defined an extremely simple toy machine language, and then a slightly more complex one to experiment with techniques of machine language manipulation.
The first machine language defined assumes that memory is just a sequence of 16-bit words and that any instruction is composed of two parts of one word each—an opcode and an operand. The second machine language assumes that memory is a sequence of bytes and some instructions can manipulate single bytes, while other instructions can manipulate whole words.
This introduced the endianness issue, which concerns how to interpret two consecutive bytes as a word. As an example, the sieve of Eratosthenes algorithm was first written in Rust and then translated into both machine languages.
For the first machine language, an interpreter was written without using any external library. It was used to first interpret a small number conversion program (word_machine_convert) and then the more complex sieve algorithm...
