/************************************************************************************************************************
 *
 * PhotoGrok Script
 *
 * Runs in multiple passes to detect duplicate files using MD5. Each tree build pass must point at a different folder(s).
 *
 * The second (or any subsequent) passes can take a long time if there are a lot of identical files. The script compares
 * files between the first pass and subsequent passes. If the name and size match, an MD5 is computed to determine if
 * the files are identical. A byte by byte comparison might be faster.
 *
 * The script fails in later versions of Java if using the Web Start version (apparently a security violation accessing
 * file.length() from JavaScript). Use the standalone Mac or PC versions instead.
 *
 * This script must be used in one of the following ways. With a filter as follows (always make it the last filter tab):
 *
 * Directory,FileName | Javascript Expression | match(Directory,FileName)   
 *
 * The filter version will return only files that have a match from the first tree build. You can also create a custom 
 * tag with, for example, two nodes: "found" and "notfound" (use the filter above as a starting point). Then build your
 * tree using that custom tag. This will take twice as long as the filter only version but could be optimized.
 *
 *************************************************************************************************************************/

// get HashMap to store data between each pass
var map = getMap();

var matches = 0;
var recorded = 0;
var md5 = null;
var hexArray = new java.lang.String("0123456789ABCDEF").toCharArray();

var init = !map.containsKey('names');

// first pass...or at least first pass since map was cleared
if(init){
	map.put('names', new java.util.HashMap());
	map.put('files', new java.util.HashMap());
	
}

function match(directory, filename){

	var file = new java.io.File(directory, filename);

	var namesMap = map.get('names');
	var filesMap = map.get('files');

	var lcfn = filename.toLowerCase();

	if(init){
		if(namesMap.containsKey(lcfn)){
			namesMap.get(lcfn).push(file.length());
			filesMap.get(lcfn).push(file);
		}else{
			var sizearray = new Array();
			var filearray = new Array();
			sizearray[0] = file.length();
			filearray[0] = file;
			namesMap.put(lcfn, sizearray);
			filesMap.put(lcfn, filearray);

		}
		recorded++;
		return true;
	}

	var match = false;

	if(namesMap.containsKey(lcfn)){
		var sizearray = namesMap.get(lcfn);
		match = sizearray.indexOf(file.length()) != -1;		
		if(match){
			var filearray = filesMap.get(lcfn);
			var f = filearray[sizearray.indexOf(file.length())];
			match = getMD5(file).equals(getMD5(f));			
		}
		
	}

	if(match){
		matches++;
	}

	return match;
}

function getMD5(file){
	var data = java.lang.reflect.Array.newInstance(java.lang.Byte.TYPE, new java.lang.Long(file.length()).intValue());
	var fis = new java.io.FileInputStream(file);
	fis.read(data);
	fis.close();

	if(md5 == null){
		md5 = java.security.MessageDigest.getInstance("MD5");
	}

	return bytesToHex(md5.digest(data));	
}

function bytesToHex(b){

	var hexChars = java.lang.reflect.Array.newInstance(java.lang.Character.TYPE, b.length * 2);
	for(var j = 0; j < b.length; j++){
		var v = b[j] & 0xFF;
		hexChars[j *2] = hexArray[v >>> 4];
		hexChars[j * 2 + 1] = hexArray[v & 0x0F];
	}
	var res = new java.lang.String(hexChars);
	return res;
}


// called when tree build is complete
function done(){
	if(init){
		showMessage('Recorded ' + recorded + ' files');
	}else{
		showMessage('Found ' + matches + ' matches');
	}

}