encode.js

var inverseXML = getInverseObj(require("../maps/xml.json")),
    xmlReplacer = getInverseReplacer(inverseXML);

exports.XML = getInverse(inverseXML, xmlReplacer);

var inverseHTML = getInverseObj(require("../maps/entities.json")),
    htmlReplacer = getInverseReplacer(inverseHTML);

exports.HTML = getInverse(inverseHTML, htmlReplacer);

function getInverseObj(obj){
	return Object.keys(obj).sort().reduce(function(inverse, name){
		inverse[obj[name]] = "&" + name + ";";
		return inverse;
	}, {});
}

function getInverseReplacer(inverse){
	return new RegExp("\\" + Object.keys(inverse).sort().join("|\\"), "g");
}

var re_nonASCII = /[^\0-\x7F]/g,
    re_astralSymbols = /[\uD800-\uDBFF][\uDC00-\uDFFF]/g;

function nonUTF8Replacer(c){
	return "&#x" + c.charCodeAt(0).toString(16).toUpperCase() + ";";
}

function astralReplacer(c){
	// http://mathiasbynens.be/notes/javascript-encoding#surrogate-formulae
	var high = c.charCodeAt(0);
	var low  = c.charCodeAt(1);
	var codePoint = (high - 0xD800) * 0x400 + low - 0xDC00 + 0x10000;
	return "&#x" + codePoint.toString(16).toUpperCase() + ";";
}

function getInverse(inverse, re){
	function func(name){
		return inverse[name];
	}

	return function(data){
		return data
				.replace(re, func)
				.replace(re_astralSymbols, astralReplacer)
				.replace(re_nonASCII, nonUTF8Replacer);
	};
}