| 1 | use crate::instructions::Instruction; |
| 2 | use std::collections::HashMap; |
| 3 | |
| 4 | fn tokenize(line: &str) -> Vec<String> { |
| 5 | line.split(|c| c == ' ' || c == ',' || c == '\t') |
| 6 | .filter(|s| !s.is_empty()) |
| 7 | .map(|s| s.to_lowercase()) |
| 8 | .collect() |
| 9 | } |
| 10 | |
| 11 | fn parse_reg(s: &str) -> u8 { |
| 12 | match s { |
| 13 | "a" => 0, |
| 14 | "b" => 1, |
| 15 | "c" => 2, |
| 16 | "d" => 3, |
| 17 | _ => panic!("Unknown register {}", s), |
| 18 | } |
| 19 | } |
| 20 | |
| 21 | fn instr_size(tokens: &[String]) -> u16 { |
| 22 | match tokens[0].as_str() { |
| 23 | "mov" | "add" | "sub" | "jmp" | "jz" | "jnz" => 3, |
| 24 | "hlt" => 1, |
| 25 | _ => panic!("Unknown instruction {}", tokens[0]) |
| 26 | |
| 27 | } |
| 28 | } |
| 29 | |
| 30 | |
| 31 | fn first_pass(source: &str) -> HashMap<String, u16> { |
| 32 | let mut symbols = HashMap::new(); |
| 33 | let mut addr: u16 = 0; |
| 34 | |
| 35 | for line in source.lines() { |
| 36 | let line = line.trim(); |
| 37 | |
| 38 | // Ignoring comments and empty lines |
| 39 | if line.is_empty() || line.starts_with(";") { |
| 40 | continue; |
| 41 | } |
| 42 | |
| 43 | // Labels (ends with ":") |
| 44 | if line.ends_with(":") { |
| 45 | let label = line.trim_end_matches(":").to_string(); |
| 46 | symbols.insert(label, addr); |
| 47 | continue; |
| 48 | } |
| 49 | |
| 50 | let tokens = tokenize(line); |
| 51 | addr += instr_size(&tokens); |
| 52 | } |
| 53 | |
| 54 | symbols |
| 55 | } |
| 56 | |
| 57 | pub fn assembler(source: &str) -> Vec<u8> { |
| 58 | let symbols = first_pass(source); |
| 59 | let mut bytes = Vec::new(); |
| 60 | |
| 61 | for (line_no, line) in source.lines().enumerate() { |
| 62 | let line = line.trim(); |
| 63 | |
| 64 | // Comments in assembly start with ";" |
| 65 | if line.is_empty() || line.starts_with(';') || line.ends_with(":") { |
| 66 | continue; |
| 67 | } |
| 68 | |
| 69 | let tokens = tokenize(line); |
| 70 | |
| 71 | match tokens[0].as_str() { |
| 72 | "mov" => { |
| 73 | // mov reg, imm |
| 74 | let reg = parse_reg(&tokens[1]); |
| 75 | let imm: u8 = tokens[2].parse().unwrap(); |
| 76 | |
| 77 | bytes.push(Instruction::MOV as u8); |
| 78 | bytes.push(reg); |
| 79 | bytes.push(imm); |
| 80 | } |
| 81 | |
| 82 | "add" => { |
| 83 | // add a, b |
| 84 | let r1 = parse_reg(&tokens[1]); |
| 85 | let r2 = parse_reg(&tokens[2]); |
| 86 | |
| 87 | bytes.push(Instruction::ADD as u8); |
| 88 | bytes.push(r1); |
| 89 | bytes.push(r2); |
| 90 | } |
| 91 | |
| 92 | "sub" => { |
| 93 | // sub a, b |
| 94 | let r1 = parse_reg(&tokens[1]); |
| 95 | let r2 = parse_reg(&tokens[2]); |
| 96 | |
| 97 | bytes.push(Instruction::SUB as u8); |
| 98 | bytes.push(r1); |
| 99 | bytes.push(r2); |
| 100 | } |
| 101 | |
| 102 | "jmp" | "jz" | "jnz" => { |
| 103 | let opcode = match tokens[0].as_str() { |
| 104 | "jmp" => Instruction::JMP, |
| 105 | "jz" => Instruction::JZ, |
| 106 | "jnz" => Instruction::JNZ, |
| 107 | _ => unreachable!() |
| 108 | }; |
| 109 | |
| 110 | let label = &tokens[1]; |
| 111 | let addr = *symbols.get(label).expect("Uknown label"); |
| 112 | |
| 113 | bytes.push(opcode as u8); |
| 114 | bytes.push((addr & 0xFF) as u8); // low |
| 115 | bytes.push((addr >> 8) as u8); // high |
| 116 | } |
| 117 | |
| 118 | "hlt" => { |
| 119 | bytes.push(Instruction::HLT as u8); |
| 120 | } |
| 121 | |
| 122 | _ => panic!("Line {}: unknown instruction", line_no + 1), |
| 123 | } |
| 124 | } |
| 125 | |
| 126 | bytes |
| 127 | } |