libcore-tokenizer

NFA Tokenizer

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-tokenizer --save

Usage

The following lines defines tokens and regular expressions and uses the tokenizer instance to tokenize the string.

var Tokenizer = require("libcore-tokenizer"),
	tokenizer = new Tokenizer(),
	subject = "12345abc67890";

// declare tokens
tokenizer.define([
	"number",	 /[0-9]+/,
        
	"sequence",	/a-c]+/,
					/[d-z]+/,
					/[A-Z]+/
]);

// tokenize (indexFrom, stringSubject)

tokenizer.tokenize(0, subject);
// result:
// 		["number", "12345", 5]


tokenizer.tokenize(5, subject);
// result: ["sequence", "abc", 8]

tokenizer.tokenize(8, subject);
// result: ["number", "67890", 13]


tokenizer.tokenize(13, subject);
// result: ["$", "", 14] -- end token, next call on next index would result in null

tokenizer.tokenize(14, subject);
// result: null

You can also export state matrix of the state machine to JSON string by running the lines below.

var Tokenizer = require("libcore-tokenizer"),
	tokenizer = new Tokenizer();

// declare tokens
tokenizer.define([
	"number",	 /[0-9]+/,
        
	"sequence",	/a-c]+/,
					/[d-z]+/,
					/[A-Z]+/
]);

console.log(tokenizer.toJSON());
// result: JSON string containing state matrix object

You can import state matrix from JSON string or from Native Object.

var Tokenizer = require("libcore-tokenizer"),
	exportedJSON = require("./tokenizer-matrix.json"),
	tokenizer = new Tokenizer();
	
tokenizer.fromJSON(exportedJSON());

// torkenizer is now ready for tokenizing string

Supported Regex Operators for Regular Expression patterns

• Alternation | operator
• Optional ? operator
• Repeat + operator
• Kleen Star * operator
• Character Class [], and Negated Character Class [^] operators. Character Ranges like 0-9a-z is also supported.
• Grouping () operator (Capture is not yet implemented)