nand2tetris/compiler/backend/lib/lexer.mll

{
open Parser
}

let period     = '.'
let colon      = ':'
let underscore = '_'
let numbers    = ['0'-'9']+
let alphabets  = ['a'-'z' 'A'-'Z']+
let white      = [' ' '\t']+
let comment    = "//" _*


let address    = (numbers)+
let id         = (alphabets | numbers | colon | period | underscore)+

rule read =
  parse
  | white          { read lexbuf }
  | comment        { read lexbuf }
  | "add"          { ADD } 
  | "sub"          { SUB }
  | "neg"          { NEG }
  | "eq"           { EQ }
  | "gt"           { GT }
  | "lt"           { LT }
  | "and"          { AND }
  | "or"           { OR }
  | "not"          { NOT }
  | "pop"          { POP }
  | "push"         { PUSH }  
  | "argument"     { ARGUMENT }  
  | "local"        { LOCAL }  
  | "static"       { STATIC }  
  | "this"         { THIS }  
  | "that"         { THAT }  
  | "pointer"      { POINTER }  
  | "temp"         { TEMP }  
  | "constant"     { CONSTANT }  
  | "label"        { LABEL }
  | "goto"         { GOTO }
  | "if-goto"      { IFGOTO }
  | "function"     { FUNCTION }
  | "call"         { CALL }
  | "return"       { RETURN }
  | address as a   { ADDRESS (int_of_string a) }
  | id as i        { ID (i) }
  | eof            { EOF }
  | _ as c         { failwith (Printf.sprintf "unexpected character: %C" c) }