| 1 | use crate::instructions::Instruction; |
| 2 | use std::collections::HashMap; |
| 3 | |
| 4 | fn tokenize(line: &str) -> Vec<String> { |
| 5 | line.split([' ', ',', '\t']) |
| 6 | .filter(|s| !s.is_empty()) |
| 7 | .map(|s| s.to_lowercase()) |
| 8 | .collect() |
| 9 | } |
| 10 | |
| 11 | // Returns corrensponding u8 value of the register |
| 12 | fn parse_reg(s: &str) -> u8 { |
| 13 | match s { |
| 14 | "a" => 0, |
| 15 | "b" => 1, |
| 16 | "c" => 2, |
| 17 | "d" => 3, |
| 18 | _ => panic!("Unknown register {}", s), |
| 19 | } |
| 20 | } |
| 21 | |
| 22 | fn is_reg(s: &str) -> bool { |
| 23 | matches!(s, "a" | "b" | "c" | "d") |
| 24 | } |
| 25 | |
| 26 | fn instr_size(tokens: &[String]) -> u16 { |
| 27 | match tokens[0].as_str() { |
| 28 | "mov" | "add" | "sub" | "jmp" | "jz" | "jnz" | "cmp" => 3, |
| 29 | "hlt" => 1, |
| 30 | _ => panic!("Unknown instruction {}", tokens[0]), |
| 31 | } |
| 32 | } |
| 33 | |
| 34 | fn first_pass(source: &str) -> HashMap<String, u16> { |
| 35 | let mut symbols = HashMap::new(); |
| 36 | let mut addr: u16 = 0; |
| 37 | |
| 38 | for line in source.lines() { |
| 39 | let line = line.trim(); |
| 40 | |
| 41 | // Ignoring comments and empty lines |
| 42 | if line.is_empty() || line.starts_with(";") { |
| 43 | continue; |
| 44 | } |
| 45 | |
| 46 | // Labels (ends with ":") |
| 47 | if line.ends_with(":") { |
| 48 | let label = line.trim_end_matches(":").to_string(); |
| 49 | symbols.insert(label, addr); |
| 50 | continue; |
| 51 | } |
| 52 | |
| 53 | let tokens = tokenize(line); |
| 54 | addr += instr_size(&tokens); |
| 55 | } |
| 56 | |
| 57 | symbols |
| 58 | } |
| 59 | |
| 60 | pub fn assembler(source: &str) -> Vec<u8> { |
| 61 | let symbols = first_pass(source); |
| 62 | let mut bytes = Vec::new(); |
| 63 | |
| 64 | for (line_no, line) in source.lines().enumerate() { |
| 65 | let line = line.trim(); |
| 66 | |
| 67 | // Comments in assembly start with ";" |
| 68 | if line.is_empty() || line.starts_with(';') || line.ends_with(":") { |
| 69 | continue; |
| 70 | } |
| 71 | |
| 72 | let tokens = tokenize(line); |
| 73 | |
| 74 | match tokens[0].as_str() { |
| 75 | "mov" => { |
| 76 | // mov reg, imm |
| 77 | let r1= parse_reg(&tokens[1]); |
| 78 | if is_reg(&tokens[2]) { |
| 79 | let r2 = parse_reg(&tokens[2]); |
| 80 | bytes.push(Instruction::MOV_RR as u8); |
| 81 | bytes.push(r1); |
| 82 | bytes.push(r2); |
| 83 | } else { |
| 84 | let imm: u8 = tokens[2].parse().unwrap(); |
| 85 | bytes.push(Instruction::MOV_RI as u8); |
| 86 | bytes.push(r1); |
| 87 | bytes.push(imm); |
| 88 | } |
| 89 | } |
| 90 | |
| 91 | "add" => { |
| 92 | // add a, b |
| 93 | let r1 = parse_reg(&tokens[1]); |
| 94 | |
| 95 | if is_reg(&tokens[2]) { |
| 96 | let r2 = parse_reg(&tokens[2]); |
| 97 | bytes.push(Instruction::ADD_RR as u8); |
| 98 | bytes.push(r1); |
| 99 | bytes.push(r2); |
| 100 | } else { |
| 101 | let imm: u8 = tokens[2].parse().unwrap(); |
| 102 | bytes.push(Instruction::ADD_RI as u8); |
| 103 | bytes.push(r1); |
| 104 | bytes.push(imm); |
| 105 | } |
| 106 | |
| 107 | } |
| 108 | |
| 109 | "sub" => { |
| 110 | // sub a, b |
| 111 | let r1 = parse_reg(&tokens[1]); |
| 112 | if is_reg(&tokens[2]) { |
| 113 | let r2 = parse_reg(&tokens[2]); |
| 114 | bytes.push(Instruction::SUB_RR as u8); |
| 115 | bytes.push(r1); |
| 116 | bytes.push(r2); |
| 117 | } else { |
| 118 | let imm: u8 = tokens[2].parse().unwrap(); |
| 119 | bytes.push(Instruction::SUB_RI as u8); |
| 120 | bytes.push(r1); |
| 121 | bytes.push(imm); |
| 122 | } |
| 123 | } |
| 124 | |
| 125 | "jmp" | "jz" | "jnz" => { |
| 126 | let opcode = match tokens[0].as_str() { |
| 127 | "jmp" => Instruction::JMP, |
| 128 | "jz" => Instruction::JZ, |
| 129 | "jnz" => Instruction::JNZ, |
| 130 | _ => unreachable!(), |
| 131 | }; |
| 132 | |
| 133 | let label = &tokens[1]; |
| 134 | let addr = *symbols.get(label).expect("Uknown label"); |
| 135 | |
| 136 | bytes.push(opcode as u8); |
| 137 | bytes.push((addr & 0xFF) as u8); // low |
| 138 | bytes.push((addr >> 8) as u8); // high |
| 139 | } |
| 140 | |
| 141 | "cmp" => { |
| 142 | let r1 = parse_reg(&tokens[1]); |
| 143 | if tokens[2].chars().all(|c| c.is_ascii_digit()) { |
| 144 | let imm: u8 = tokens[2].parse().unwrap(); |
| 145 | bytes.push(Instruction::CMP_RI as u8); |
| 146 | bytes.push(r1); |
| 147 | bytes.push(imm); |
| 148 | } else { |
| 149 | let r2 = parse_reg(&tokens[2]); |
| 150 | bytes.push(Instruction::CMP_RR as u8); |
| 151 | bytes.push(r1); |
| 152 | bytes.push(r2); |
| 153 | } |
| 154 | } |
| 155 | |
| 156 | "hlt" => { |
| 157 | bytes.push(Instruction::HLT as u8); |
| 158 | } |
| 159 | |
| 160 | _ => panic!("Line {}: unknown instruction", line_no + 1), |
| 161 | } |
| 162 | } |
| 163 | |
| 164 | bytes |
| 165 | } |