Non-incremental parser for WebVTT files.

WebVTTParser.js

/*
http://www.whatwg.org/specs/web-apps/current-work/webvtt.html
*/
function parseWebVTT(input){
	"use strict";
	var line,l,p,cue_list=[],
		cue,cue_text,id,fields,
		time_pat = /\s*(\d*:?[0-5]\d:[0-5]\d\.\d\d\d)\s*-->\s*(\d*:?[0-5]\d:[0-5]\d\.\d\d\d)\s*(.*)/;

	//If the first character is a BYTE ORDER MARK (BOM) character, advance position to the next character in input.
	l = p = +(input[0] === '\uFEFF');
	//Collect a sequence of characters that are not CR or LF characters.
	while(p<input.length && input[p]!=='\r' && input[p] !=='\n'){p++;}
	//If line is less than six characters long, then abort these steps. The file is not a WebVTT file.
	if(p-l<6){throw new Error("Not WebVTT Data");}
	line = input.substring(l,p);
	//If the first six characters do not exactly equal "WEBVTT", then abort these steps. The file is not a WebVTT file.
	//If line is more than six characters long but the seventh character is neither a U+0020 SPACE character nor a U+0009 CHARACTER TABULATION (tab) character, then abort these steps. The file is not a WebVTT file.
	if(!/^WEBVTT([\u0020\u0009].*|$)/.test(line)){throw new Error("Not WebVTT Data");}
	//If position is past the end of input, then jump to the step labeled end.
	if(p>=input.length){return cue_list;}
	do{	//Header:
		if(input[p] === '\r'){ //Skip CR
			//If position is past the end of input, then jump to the step labeled end.
			if(++p>=input.length){return cue_list;}
		}
		if(input[p] === '\n'){//Skip LF
			if(++p>=input.length){return cue_list;}
		}
		l=p; //Collect a sequence of characters that are not CR or LF characters.
		while(input[p] !== '\r' && input[p] !== '\n'){
			if(++p>=input.length){return cue_list;}
		}
	}while(l!==p);	//If line is not the empty string, then jump back to the step labeled header.
	
	cue_loop: do{
		//Skip CR & LF characters.
		while(input[p]==='\r' || input[p]==='\n'){
			if(++p>=input.length){break cue_loop;}
		}
		l=p; //Collect a sequence of characters that are not CR or LF characters.
		while(input[p]!=='\r' && input[p] !=='\n'){
			if(++p>=input.length){break cue_loop;}
		}
		line = input.substring(l,p);
		//If line does not contain "-->", treat it as an id & get a new line
		if(line.indexOf('-->')===-1){
			if(input[p] === '\r'){ //Skip CR
				if(++p>=input.length){break cue_loop;}
			}
			if(input[p] === '\n'){ //Skip LF
				if(++p>=input.length){break cue_loop;}
			}
			l=p; //Collect a sequence of characters that are not CR or LF characters.
			while(input[p]!=='\r' && input[p] !=='\n'){
				if(++p>=input.length){break cue_loop;}
			} //If line is the empty string, jump to the step labeled cue loop.
			if(l===p){continue cue_loop;}
			id = line; //Let cue's text track cue identifier be the previous line.
			line = input.substring(l,p);
		}else{id = '';}
		cue = { //set default cue parameters
			id:id,				text:'',
			pause_on_exit:false,
			wdir:'horizontal',	snap:true,
			line:'auto',		position:50,
			size:100,			align:'middle'
		};
		//Timings:
		try{ //Collect WebVTT cue timings and settings from line, using cue for the results.
			if(!(fields = time_pat.exec(line))){throw new Error("Invalid Timestamp Data");}
			cue.start = parse_timestamp(fields[1]);
			cue.stop = parse_timestamp(fields[2]);
			parse_settings(cue,fields[3]);
		} catch(e) {
			console.log(e.stack);
			do{	//Bad cue loop:
				if(input[p] === '\r'){ //Skip CR
					if(++p>=input.length){break cue_loop;}
				}
				if(input[p] === '\n'){ //Skip LF
					if(++p>=input.length){break cue_loop;}
				}
				l=p; //Collect a sequence of characters that are not CR or LF characters.
				while(input[p]!=='\r' && input[p] !=='\n'){
					if(++p>=input.length){break cue_loop;}
				}
			}while(l!==p);//If line is the empty string, then jump to the step labeled cue loop.
			continue cue_loop;
		}
		cue_text = [];
		do{	//Cue text loop:
			if(input[p] === '\r'){ //Skip CR
				//If position is past the end of input, jump to cue text processing.
				if(++p===input.length){break;}
			}
			if(input[p] === '\n'){ //Skip LF
				if(++p===input.length){break;}
			}
			l=p; //Collect a sequence of characters that are not CR or LF characters.
			while(p<input.length && input[p]!=='\r' && input[p] !=='\n'){p++;}
			if(l===p){break;}	//If line is the empty string, then jump to the step labeled cue text processing.
			//Replace all U+0000 NULL characters in input by U+FFFD REPLACEMENT CHARACTERs.
			cue_text.push(input.substring(l,p).replace('\0','\uFFFD'));
		}while(p<input.length);
		//Cue text processing:
		cue.text = cue_text.join('\n');
		//This where we ought to construct the cue-text DOM
		cue_list.push(cue);	//Add cue to the text track list of cues output.
	}while(p<input.length);
	//End: The file has ended. The WebVTT parser has finished.
	return cue_list;
};

function parse_timestamp(input){
	var ret,p,fields;
	"use strict";
	if(input[0]===':'){throw new SyntaxError("Unexpected Colon");}
	fields = input.split(/[:.]/);
	if(fields.length===4){
		ret = parseInt(fields[0],10)*3600+parseInt(fields[3],10)/1000;
		p = 1;
	}else{
		ret = parseInt(fields[2],10)/1000;
		p = 0;
	}
	return ret + parseInt(fields[p],10)*60 + parseInt(fields[++p],10);
}

function parse_settings(cue,input){
	"use strict";
	var match,value,number,
		set_pat = /([ADLST]):(\S+)/g;
	while(!!(match = set_pat.exec(input))){
		value = match[2];
		switch(match[1]){
			case 'A': //Alignment
				if(value==='start' || value==='middle' || value==='end'){cue.align=value;}
				continue;
			case 'D': //Text direction
				if(value === 'vertical' || value === 'vertical-lr'){cue.wdir = value;}
				continue;
			case 'L': //Line position
				if(/^-?\d+%?$/.test(value)){
					number = parseInt(value,10);
					if(value[value.length-1] === '%'){	//If the last character in value is %
						if(number<0 || number>100){continue;}
						cue.snap = false;
					}
					cue.line = number;
				}
				continue;
			case 'S': //Text Size
				if(/^\d+%$/.test(value)){
					number = parseInt(value,10);
					if(number>=0 && number<=100){
						cue.size = number;
					}
				}
				continue;
			case 'T': //Text Position
				if(/^\d+%$/.test(value)){
					number = parseInt(value,10);
					if(number>=0 && number<=100){
						cue.position = number;
					}
				}
				continue;
		}
	}
}