Parser.js

var Tokenizer = require("./Tokenizer.js");

/*
	Options:

	xmlMode: Special behavior for script/style tags (true by default)
	lowerCaseAttributeNames: call .toLowerCase for each attribute name (true if xmlMode is `false`)
	lowerCaseTags: call .toLowerCase for each tag name (true if xmlMode is `false`)
*/

/*
	Callbacks:

	oncdataend,
	oncdatastart,
	onclosetag,
	oncomment,
	oncommentend,
	onerror,
	onopentag,
	onprocessinginstruction,
	onreset,
	ontext
*/

var formTags = {
	input: true,
	option: true,
	optgroup: true,
	select: true,
	button: true,
	datalist: true,
	textarea: true
};

var openImpliesClose = {
	tr      : { tr:true, th:true, td:true },
	th      : { th:true },
	td      : { thead:true, td:true },
	body    : { head:true, link:true, script:true },
	li      : { li:true },
	p       : { p:true },
	select  : formTags,
	input   : formTags,
	output  : formTags,
	button  : formTags,
	datalist: formTags,
	textarea: formTags,
	option  : { option:true },
	optgroup: { optgroup:true }
};

var voidElements = {
	__proto__: null,
	area: true,
	base: true,
	basefont: true,
	br: true,
	col: true,
	command: true,
	embed: true,
	frame: true,
	hr: true,
	img: true,
	input: true,
	isindex: true,
	keygen: true,
	link: true,
	meta: true,
	param: true,
	source: true,
	track: true,
	wbr: true
};

var re_nameEnd = /\s|\//;

function Parser(cbs, options){
	this._options = options || {};
	this._cbs = cbs || {};

	this._tagname = "";
	this._attribname = "";
	this._attribvalue = "";
	this._attribs = null;
	this._stack = [];
	this._done = false;

	this.startIndex = 0;
	this.endIndex = null;

	this._tokenizer = new Tokenizer(options, this);
}

require("util").inherits(Parser, require("events").EventEmitter);

Parser.prototype._updatePosition = function(initialOffset){
	if(this.endIndex === null){
		this.startIndex = this._tokenizer._sectionStart <= initialOffset ? 0 : this._tokenizer._sectionStart - initialOffset;
	}
	this.startIndex = this.endIndex + 1;
	this.endIndex = this._tokenizer._index;
};

//Tokenizer event handlers
Parser.prototype.ontext = function(data){
	this._updatePosition(1);
	this.endIndex--;

	if(this._cbs.ontext) this._cbs.ontext(data);
};

Parser.prototype.onopentagname = function(name){
	if(!(this._options.xmlMode || "lowerCaseTags" in this._options) || this._options.lowerCaseTags){
		name = name.toLowerCase();
	}

	this._tagname = name;

	if (!this._options.xmlMode && name in openImpliesClose) {
		for(
			var el;
			(el = this._stack[this._stack.length-1]) in openImpliesClose[name];
			this.onclosetag(el)
		);
	}

	if(this._options.xmlMode || !(name in voidElements)){
		this._stack.push(name);
	}

	if(this._cbs.onopentagname) this._cbs.onopentagname(name);
	if(this._cbs.onopentag) this._attribs = {};
};

Parser.prototype.onopentagend = function(){
	this._updatePosition(1);
    
	if(this._attribs){
		if(this._cbs.onopentag) this._cbs.onopentag(this._tagname, this._attribs);
		this._attribs = null;
	}
    
	if(!this._options.xmlMode && this._cbs.onclosetag && this._tagname in voidElements){
		this._cbs.onclosetag(this._tagname);
	}
    
	this._tagname = "";
};

Parser.prototype.onclosetag = function(name){
	this._updatePosition(1);

	if(!(this._options.xmlMode || "lowerCaseTags" in this._options) || this._options.lowerCaseTags){
		name = name.toLowerCase();
	}

	if(this._stack.length && (!(name in voidElements) || this._options.xmlMode)){
		var pos = this._stack.lastIndexOf(name);
		if(pos !== -1){
			if(this._cbs.onclosetag){
				pos = this._stack.length - pos;
				while(pos--) this._cbs.onclosetag(this._stack.pop());
			}
			else this._stack.length = pos;
		} else if(name === "p" && !this._options.xmlMode){
			this.onopentagname(name);
			this._closeCurrentTag();
		}
	} else if(!this._options.xmlMode && (name === "br" || name === "p")){
		this.onopentagname(name);
		this._closeCurrentTag();
	}
};

Parser.prototype.onselfclosingtag = function(){
	if(this._options.xmlMode){
		this._closeCurrentTag();
	} else {
		this.onopentagend();
	}
};

Parser.prototype._closeCurrentTag = function(){
	var name = this._tagname;

	this.onopentagend();

	//self-closing tags will be on the top of the stack
	//(cheaper check than in onclosetag)
	if(this._stack[this._stack.length-1] === name){
		if(this._cbs.onclosetag){
			this._cbs.onclosetag(name);
		}
		this._stack.pop();
	}
};

Parser.prototype.onattribname = function(name){
	if(!(this._options.xmlMode || "lowerCaseAttributeNames" in this._options) || this._options.lowerCaseAttributeNames){
		name = name.toLowerCase();
	}
	this._attribname = name;
};

Parser.prototype.onattribdata = function(value){
	this._attribvalue += value;
};

Parser.prototype.onattribend = function(){
	if(this._cbs.onattribute) this._cbs.onattribute(this._attribname, this._attribvalue);
	if(
		this._attribs &&
		!Object.prototype.hasOwnProperty.call(this._attribs, this._attribname)
	){
		this._attribs[this._attribname] = this._attribvalue;
	}
	this._attribname = "";
	this._attribvalue = "";
};

Parser.prototype.ondeclaration = function(value){
	if(this._cbs.onprocessinginstruction){
		var idx = value.search(re_nameEnd),
		    name = idx < 0 ? value : value.substr(0, idx);

		if(!(this._options.xmlMode || "lowerCaseTags" in this._options) || this._options.lowerCaseTags){
			name = name.toLowerCase();
		}
		this._cbs.onprocessinginstruction("!" + name, "!" + value);
	}
};

Parser.prototype.onprocessinginstruction = function(value){
	if(this._cbs.onprocessinginstruction){
		var idx = value.search(re_nameEnd),
		    name = idx < 0 ? value : value.substr(0, idx);

		if(!(this._options.xmlMode || "lowerCaseTags" in this._options) || this._options.lowerCaseTags){
			name = name.toLowerCase();
		}
		this._cbs.onprocessinginstruction("?" + name, "?" + value);
	}
};

Parser.prototype.oncomment = function(value){
	this._updatePosition(4);

	if(this._cbs.oncomment) this._cbs.oncomment(value);
	if(this._cbs.oncommentend) this._cbs.oncommentend();
};

Parser.prototype.oncdata = function(value){
	this._updatePosition(1);

	if(this._options.xmlMode){
		if(this._cbs.oncdatastart) this._cbs.oncdatastart();
		if(this._cbs.ontext) this._cbs.ontext(value);
		if(this._cbs.oncdataend) this._cbs.oncdataend();
	} else {
		this.oncomment("[CDATA[" + value + "]]");
	}
};

Parser.prototype.onerror = function(err){
	if(this._cbs.onerror) this._cbs.onerror(err);
};

Parser.prototype.onend = function(){
	if(this._cbs.onclosetag){
		for(
			var i = this._stack.length;
			i > 0;
			this._cbs.onclosetag(this._stack[--i])
		);
	}
	if(this._cbs.onend) this._cbs.onend();
};


//Resets the parser to a blank state, ready to parse a new HTML document
Parser.prototype.reset = function(){
	if(this._cbs.onreset) this._cbs.onreset();
	this._tokenizer.reset();

	this._tagname = "";
	this._attribname = "";
	this._attribs = null;
	this._stack = [];
	this._done = false;
};

//Parses a complete HTML document and pushes it to the handler
Parser.prototype.parseComplete = function(data){
	this.reset();
	this.end(data);
};

Parser.prototype.write = function(chunk){
	if(this._done) this.onerror(Error(".write() after done!"));
	this._tokenizer.write(chunk);
};

Parser.prototype.end = function(chunk){
	if(this._done) this.onerror(Error(".end() after done!"));
	this._tokenizer.end(chunk);
	this._done = true;
};

//alias for backwards compat
Parser.prototype.parseChunk = Parser.prototype.write;
Parser.prototype.done = Parser.prototype.end;

module.exports = Parser;