libcore-parser-lalr

LALR(1) Parser for LALR grammar

Installation

Tokenizer can be installed from NPM by running the lines below in your working directory containing package.json file for use in NodeJS, browserify or webpack.

npm install libcore-lalr --save

Usage

The following lines defines a parser declaring grammar rules composed of tokens (RegExp), recursion of rules, and a combination of both.

var lalr = require("libcore-parser-lalr");
var parser = lalr.define("Expr", // Root production
                [
                    // Lexical grammar
                    "+",    [
                                /\+/
                            ],

                    "*",    [
                                /\*/
                            ],

                    "(",    [/\(/],

                    ")",    [/\)/],

                    "number",   [/(\+|\-)?[0-9]+(\.[0-9]+)?/],

                    "whitespace", [/[ \r\n\t]+/],

                    "identifier", [/[a-zA-Z\_\$][a-zA-Z0-9\_\$]*/],

                    // Grammar rules
                    "Expr", [
                                "Ass"                   // 1:Expr rule
                            ],

                    "Ass",  [
                                "Add",                  // 1:Ass rule
                                                        // 2:Ass rule
                                ["identifier", /\=/, "Ass"]
                            ],

                    "Add",  [
                                "Mul",                  // Add1 rule
                                ["Add", "+", "Mul"],    // Add2 rule
                            ],
                    "Mul",  [
                                "Unit"                  // Mul1 rule
                                ["Mul", "*", "Unit"],   // Mul2 rule
                            ],
                    "Unit", [
                                "number",               // Unit1 rule
                                ["(", "Expr", ")"]      // Unit2 rule
                            ]
                    
                ],
                // ignore these tokens
                [
                        "whitespace"
                ]);

You can set parse subject to the iterator before you can parse. The following lines defines default iterator and parse the String subject.

var iterator = parser.iterator();
var lexeme;

// set string subject to parse
iterator.set('1 + 2 * 3');

// iterate
for (lexeme = iterator.next(); lexeme; lexeme = iterator.next()) {
    console.log(lexeme.name,            // grammar rule name
                lexeme.rule,            // grammar rule id (e.g. Mul1, Unit2)
                lexeme.value,           // lexeme value - you update this with lexeme.update("value")
                lexeme.reduceCount,     // number of lexemes popped to reduce
                lexeme);
}

For supported Regex operators for tokenizer, please refer to libcore-tokenizer

License

This Project is fully Open Source MIT licensed.