Last active
February 16, 2025 21:00
-
-
Save wilyJ80/ad77e94f70b4824f89885bcb17743007 to your computer and use it in GitHub Desktop.
JS class-based table-driven lexer for a simple Forth
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
import * as readline from 'node:readline/promises'; | |
import { stdin as input, stdout as output } from 'node:process'; | |
/* valid tokens: | |
numbers (consisting of digits), | |
words (consisting of alnum chars), | |
comment (ignore) | |
*/ | |
const Transition = class { | |
#nextState; | |
#callback; | |
#category; | |
#isOther; | |
constructor(nextState, callback, | |
category, isOther) { | |
this.#nextState = nextState; | |
this.#callback = callback; | |
this.#category = category; | |
this.#isOther = isOther; | |
} | |
get nextState() { | |
return this.#nextState; | |
} | |
get callback() { | |
return this.#callback; | |
} | |
get category() { | |
return this.#category; | |
} | |
get other() { | |
return this.#isOther; | |
} | |
} | |
const CharTester = class { | |
isLetter(ch) { | |
return new RegExp('[a-z]').test(ch); | |
} | |
isDigit(ch) { | |
return new RegExp('[0-9]').test(ch); | |
} | |
isNotLetterNorDigit(ch) { | |
return !( | |
new RegExp('[a-z]').test(ch) || | |
new RegExp('[0-9]').test(ch) | |
); | |
} | |
isDigitOrLetter(ch) { | |
return new RegExp('[a-z]|[0-9]').test(ch); | |
} | |
isParenOpen(ch) { | |
return new RegExp('\(').test(ch); | |
} | |
isParenClose(ch) { | |
return new RegExp('\)').test(ch); | |
} | |
isNotParenClose(ch) { | |
return new RegExp('^)').test(ch); | |
} | |
isBlank(ch) { | |
return new RegExp('\s').test(ch); | |
} | |
} | |
const Transitions = class { | |
#transitionTable; | |
#tester; | |
get table() { | |
return this.#transitionTable; | |
} | |
get tester() { | |
return this.#tester; | |
} | |
constructor() { | |
this.#tester = new CharTester(); | |
this.#transitionTable = [ | |
// State 0 | |
[ | |
new Transition(0, | |
this.#tester.isBlank, | |
'nonaccepting', | |
'notother'), | |
new Transition(1, | |
this.#tester.isDigit, | |
'nonaccepting', | |
'notother'), | |
new Transition(2, | |
this.#tester.isLetter, | |
'nonaccepting', | |
'notother'), | |
new Transition(5, | |
this.#tester.isParenOpen, | |
'nonaccepting', | |
'notother') | |
], | |
// State 1 | |
[ | |
new Transition(2, | |
this.#tester.isLetter, | |
'nonaccepting', | |
'notother'), | |
new Transition(3, | |
this.#tester.isNotLetterNorDigit, | |
'number', | |
'isother') | |
], | |
// State 2 | |
[ | |
new Transition(2, | |
this.#tester.isDigitOrLetter, | |
'nonaccepting', | |
'notother'), | |
new Transition(4, | |
this.#tester.isNotLetterNorDigit, | |
'word', | |
'isother') | |
], | |
// State 3: accepting | |
[], | |
// State 4: accepting | |
[], | |
// State 5 | |
[ | |
new Transition(5, | |
this.#tester.isNotParenClose, | |
'nonaccepting', | |
'isother'), | |
new Transition(6, | |
this.#tester.isParenClose, | |
'comment', | |
'notother') | |
], | |
// State 6: accepting | |
[] | |
]; | |
} | |
} | |
const Lexer = class { | |
#transitionTable; | |
#lineIndex; | |
#currentState; | |
constructor() { | |
this.#transitionTable = new Transitions().table; | |
this.#lineIndex = 0; | |
this.#currentState = 0; | |
} | |
get lineIndex() { | |
return this.#lineIndex; | |
} | |
get transitionTable() { | |
return this.#transitionTable; | |
} | |
/** @param {Number} state */ | |
set currentState(state) { | |
this.#currentState = state; | |
} | |
debug(line) { | |
console.log(line[this.#lineIndex]); | |
} | |
nextToken(line) { | |
/* transition has: next state, function, category, isother */ | |
let token = {}; | |
let transitionFound = false; | |
for (const transition of this.#transitionTable) { | |
const currentChar = line[this.#lineIndex]; | |
if (transition.callback(currentChar)) { | |
transitionFound = true; | |
// if transition is valid: advance state | |
this.currentState = transition.nextState; | |
// not valid: evaluate next transition function: ok, forof does it | |
// nonaccepting state, no valid transition found: | |
// throw error with offending built token so far | |
// evaluate until end of string (already happening with forof) | |
// reset state to 0 if accepting | |
// nonaccepting, transition found: keep buiding up the token | |
} | |
} | |
// remember to reset the charIndex later | |
// after lexing the whole current line | |
} | |
} | |
const lexer = new Lexer(); | |
const rl = readline.createInterface({ input, output }); | |
console.log('Enter q to quit'); | |
const loop = async () => { | |
const answer = await rl.question('VF # '); | |
if (answer === 'q') { | |
rl.close(); | |
} else { | |
lexer.debug(answer); | |
if (answer !== '') { | |
console.log(''); | |
} | |
loop(); | |
} | |
} | |
loop(); | |
// TODO |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment