GitHub release Build Status Coverage Status

A small tool for calculating first, follow and predict sets for the grammar. Its size is only 390 bytes (minified and gzipped). No dependencies. Size Limit controls the size.

$ npm install first-follow

You can try this tool here.

const firstFollow = require('first-follow');
const rules = [
 // S -> a b A
 {
 left: 'S',
 right: ['a', 'b', 'A']
 },
 // A -> b c
 {
 left: 'A',
 right: ['b', 'c']
 },
 // A -> ε
 {
 left: 'A',
 right: [null]
 }
];
const { firstSets, followSets, predictSets } = firstFollow(rules);
console.log(firstSets);
/*
 * // S: a
 * // A: b, ε
 *
 * {
 * S: ['a'],
 * A: ['b', null]
 * }
 */
console.log(followSets);
/*
 * // S: ┤
 * // A: ┤
 *
 * {
 * S: ['\u0000'],
 * A: ['\u0000']
 * }
 */
console.log(predictSets);
/*
 * // 1: a
 * // 2: b
 * // 3: ┤
 * 
 * {
 * '1': ['a'],
 * '2': ['b'],
 * '3': ['\u0000']
 * }
*/

The grammar is represented by array of objects. Each object describes the only one rule. The rule's object contains two required fields:

• left — specifies the left part of the rule. A single nonterminal: A, B, Program, Expression, etc.
• right — specifies the right part of the rule. It contains terminals and nonterminals or empty chain (epsilon): A + B, d * A, ε, etc.

• The firstSets object's keys are nonterminals and values are first sets for these nonterminals.
• The followSets object's keys are nonterminals and values are follow sets for these nonterminals.
• The predictSets object's keys are rules numbers (starting from 1) and values are predict sets for these rules.

• An empty chain (ε) is represented by null.
• An end mark (┤) is represented by \u0000.