icondumper2/icon.js

656 lines
23 KiB
JavaScript

//todo: Make this a module/mjs file. C6 compatibility can stay, if needed.
//LOOKING FOR: LZARI implementation (for MAX), description of CBS compression (node zlib doesn't tackle it, even with RC4'ing the data)
var ICONJS_DEBUG = false;
var ICONJS_STRICT = true;
/**
* The current version of the library.
* @constant {string}
* @default
*/
const ICONJS_VERSION = "0.6.1+u1";
/**
* Extension of DataView to add shortcuts for datatypes that I use often.
* @augments DataView
* @constructor
* @param {ArrayBuffer} buffer ArrayBuffer to base DataView from.
* @returns {Object.<string, function(number): number>} [u16le, f16le, u32le, f32le]
* @returns {Object.<string, function(number): Object.<string.number>>} [t64le]
* @access protected
*/
class yellowDataReader extends DataView {
/** Unsigned 16-bit, Little Endian.
* @param {number} i Indice offset.
* @returns {number}
*/
u16le(i){return super.getUint16(i, 1)};
/** Fixed-point 16-bit, Little Endian.
* @param {number} i Indice offset.
* @returns {number}
*/
f16le(i){return (super.getInt16(i, 1) / 4096)};
/** Unsigned 32-bit, Little Endian.
* @param {number} i Indice offset.
* @returns {number}
*/
u32le(i){return super.getUint32(i, 1)};
/** Floating-point 32-bit, Little Endian.
* @param {number} i Indice offset.
* @returns {number}
*/
f32le(i){return super.getFloat32(i, 1)};
/** 64-bit Timestamp structure, Little Endian.
* Time returned is set for JST (UTC+09:00) instead of UTC.
* Time returned is going to be offseted for JST (GMT+09:00).
* @param {number} i Indice offset.
* @returns {Object.<string, number>}
* @property {number} seconds - Seconds.
* @property {number} minutes - Minutes.
* @property {number} hours - Hours.
* @property {number} day - Day.
* @property {number} month - Month.
* @property {number} year - Year.
*/
t64le(i){return {
seconds: super.getUint8(i+1),
minutes: super.getUint8(i+2),
hours: super.getUint8(i+3),
day: super.getUint8(i+4),
month: super.getUint8(i+5),
year: super.getUint16(i+6, 1)
}};
constructor(buffer) {
super(buffer);
return {
u16le: this.u16le.bind(this),
f16le: this.f16le.bind(this),
u32le: this.u32le.bind(this),
f32le: this.f32le.bind(this),
t64le: this.t64le.bind(this)
}
}
}
/**
* Enable or disable use of debugging information in console via console.debug()
* @param {boolean} value - Enable/disable this feature
* @default false
* @public
*/
function setDebug(value) {
ICONJS_DEBUG = !!value;
}
/**
* Select if invalid characters in titles should be replaced with either spaces or nulls
* @param {boolean} value - true: with nulls, false: with spaces
* @default true
* @deprecated Hasn't been needed for a while. Dropping support by ESM transition.
* @public
*/
function setStrictness(value) {
ICONJS_STRICT = !!value;
}
/**
* Converts a texture format to a generalized texture type character.
* @param {number} i - texture format
* @returns {string} U: uncompressed, N: none, C: compressed
* @access protected
*/
function getTextureFormat(i) {
if (i<8) {
if(i==3) {
return 'N';
}
return 'U';
} else if (i>=8) {
return 'C';
} else {
return void(0);
}
}
/**
* Decompress a compressed texture using RLE.
* @param {ArrayBuffer} input - texture
* @returns {!Uint16Array} decompressed texture, equivalent to an uncompressed texture
* @public
*/
function uncompressTexture(texData) {
// for texture formats 8-15
if (texData.length & 1) {
throw "Texture size isn't a multiple of 2 (was ${texData.length})";
}
const {u16le} = new yellowDataReader(texData);
let uncompressed = new Uint16Array(16384);
let offset = 0;
for (let index = 0; index < 16384;) {
currentValue = u16le(offset);
if(currentValue === 0) {
// if this is specifically a katamari 1 or 2 icon, skip this byte
// because it's formatted like that for some reason
offset += 2;
currentValue = u16le(offset);
}
offset += 2;
if (currentValue >= 0xff00) {
//do a raw copy of the next currentValue bytes
let length = ((0x10000 - currentValue));
for (let enumerator = 0; enumerator < length; enumerator++) {
uncompressed[index] = u16le(offset);
offset += 2;
index++;
}
} else {
//repeat next byte currentValue times
uncompressed[index] = u16le(offset);
for (let indey = 0; indey < currentValue; indey++) {
uncompressed[index] = u16le(offset);
index++
}
offset += 2;
}
}
return uncompressed;
}
/**
* Converts a BGR5A1 texture to a RGB5A1 texture
* @param {Uint16Array} bgrData - texture
* @returns {!Uint16Array} converted texture
* @public
*/
function convertBGR5A1toRGB5A1(bgrData) {
if(bgrData.byteLength !== 32768) {
throw `Not a 128x128x16 texture. (length was ${bgrData.length})`;
}
// converts 5-bit blue, green, red (in that order) with one alpha bit to GL-compatible RGB5A1
let converted = new Uint16Array(16384);
for (let index = 0; index < 16384; index++) {
let b = ( bgrData[index] & 0b11111);
let g = ((bgrData[index] >> 5) & 0b11111);
let r = ((bgrData[index] >> 10) & 0b11111);
// rrrrrgggggbbbbba (a = 1 because 0 is 127 which is 1.0f opacity for the GS)
let newValue = 0b0000000000000001;
// mind you, the alpha bit ^ seems to be set randomly on textures, anyway. Maybe i'm reading these wrong?
newValue |= (r << 1);
newValue |= (g << 6);
newValue |= (b << 11);
converted[index] = newValue;
}
return converted;
}
/**
* Takes a string and returns the first part before a null character.
* @param {string} dirty - String to slice from first null character.
* @returns {string} Substring of dirty before first null character.
* @access protected
*/
function stringScrubber(dirty) {
return dirty.replace(/\0/g, "").substring(0, (dirty.indexOf("\x00") === -1) ? dirty.length : dirty.indexOf("\x00"));
}
/**
* Read a PS2D format file (icon.sys)
* @param {ArrayBuffer} input - icon.sys formatted file
* @returns {Object} (user didn't write a description)
* @public
*/
function readPS2D(input) {
const {u32le, f32le} = new yellowDataReader(input);
//!pattern ps2d.hexpat
const header = u32le(0);
if (header !== 0x44325350) {
throw `Not a PS2D file (was ${header}, expected ${0x44325350})`;
}
//:skip 2
const titleOffset = u32le(6);
//:skip 2
const bgAlpha = u32le(12); // should read as a u8, (k/127?.5) for float value (255 = a(2.0~))
const bgColors = [
{r: u32le(16), g: u32le(20), b: u32le(24), a: u32le(28)},
{r: u32le(32), g: u32le(36), b: u32le(40), a: u32le(44)},
{r: u32le(48), g: u32le(52), b: u32le(56), a: u32le(60)},
{r: u32le(64), g: u32le(68), b: u32le(72), a: u32le(76)}
]; // top-left, top-right, bottom-left, bottom-right;
const lightIndices = [
{x: f32le(80), y: f32le(84), z: f32le(88)}, //:skip 4
{x: f32le(96), y: f32le(100), z: f32le(104)}, //:skip 4
{x: f32le(112), y: f32le(116), z: f32le(120)} //:skip 4
]
const lightColors = [
{r: f32le(128), g: f32le(132), b: f32le(136), a: f32le(140)},
{r: f32le(144), g: f32le(148), b: f32le(152), a: f32le(156)},
{r: f32le(160), g: f32le(164), b: f32le(168), a: f32le(172)},
{r: f32le(176), g: f32le(180), b: f32le(184), a: f32le(188)}
]
// P2SB says color 1 is ambient, 2-4 are for 3-point cameras
// official HDD icon.sys files (completely different PS2ICON text-based format) also say the same.
const int_title = input.slice(0xc0, 0x100);
const tmp_title16 = new Uint16Array(int_title);
for (let index = 0; index < 32; index++) {
//find "bad" shift-jis two-bytes, and convert to spaces (or NULL if strict)
if(tmp_title16[index] === 129 || tmp_title16[index] === 33343) {
console.warn(`PS2D syntax error: known bad two-byte sequence 0x${tmp_title16[index].toString(16).padEnd(4,"0")} (at ${index}). Replacing with ${(ICONJS_STRICT) ? "NULL" : "0x8140"}.\n${(ICONJS_STRICT) ? "This is console-accurate, so" : "An actual console does not do this."} I recommend patching your icon.sys files!`);
tmp_title16[index] = (ICONJS_STRICT) ? 0 : 0x4081; // is a reference, so this'll edit int_title too
}
}
//:skip 4 -- Unless proven, keep 64 bytes for display name.
const int_filename_n = input.slice(0x104, 0x143);
const int_filename_c = input.slice(0x144, 0x183);
const int_filename_d = input.slice(0x184, 0x1C3);
//;skip 512 -- rest of this is just padding
const rawTitle = (new TextDecoder("shift-jis")).decode(int_title);
const title = [
stringScrubber(rawTitle.substring(0,(titleOffset/2))),
(rawTitle.indexOf("\x00") < (titleOffset/2)) ? "" : stringScrubber(rawTitle.substring((titleOffset/2)))
];
const filenames = {
n: stringScrubber((new TextDecoder("utf-8")).decode(int_filename_n)),
c: stringScrubber((new TextDecoder("utf-8")).decode(int_filename_c)),
d: stringScrubber((new TextDecoder("utf-8")).decode(int_filename_d))
}
if(ICONJS_DEBUG){
console.debug({header, titleOffset, bgAlpha, bgColors, lightIndices, lightColors, title, filenames});
}
return {filenames, title, background: {colors: bgColors, alpha: bgAlpha}, lighting: {points: lightIndices, colors: lightColors}};
}
/**
* Read a Model file ({*}, usually ICN or ICO, however)
* @param {ArrayBuffer} input - icon model file
* @returns {Object} (user didn't write a description)
* @public
*/
function readIconFile(input) {
//!pattern ps2icon-hacked.hexpat
const {u32le, f32le, f16le} = new yellowDataReader(input);
const u32_rgba8 = function(i) {return {
r: (i & 0xff),
g: ((i & 0xff00) >> 8),
b: ((i & 0xff0000) >> 16),
a: (i > 0x7fffffff ? 255 : (((i & 0xff000000) >>> 24) * 2)+1)
// I don't think alpha transparency is actually USED in icons?, rendering with it looks strange.
}};
const magic = u32le(0);
if (magic !== 0x010000) {
// USER NOTICE: So far, I have yet to parse an icon that hasn't had 0x00010000 as it's magic.
// Can someone provide me pointers to such an icon if one exists?
throw `Not a PS2 icon file (was ${magic}, expected ${0x010000})`;
}
const numberOfShapes = u32le(4);
if(numberOfShapes > 16) {
throw `Too many defined shapes! Is this a valid file? (file reports ${numberOfShapes} shapes)`;
}
const textureType = u32le(8);
const textureFormat = getTextureFormat(textureType);
//:skip 4
const numberOfVertexes = u32le(16);
if(!!(numberOfVertexes % 3)){
throw `Not enough vertices to define a triangle (${numberOfVertexes % 3} vertices remained).`;
}
// format: [xxyyzzaa * numberOfShapes][xxyyzzaa][uuvvrgba], ((8 * numberOfShapes) + 16) [per chunk]
let offset = 20;
let vertices = new Array();
const chunkLength = ((numberOfShapes * 8) + 16);
// for numberOfVertexes, copy chunkLength x times, then append with normal, uv and rgba8 color. Floats are 16-bit.
for (let index = 0; index < numberOfVertexes; index++) {
let shapes = new Array();
for (let indey = 0; indey < numberOfShapes; indey++) {
shapes.push({
x: f16le(offset+((chunkLength*index)+(indey*8))),
y: f16le(offset+((chunkLength*index)+(indey*8))+2),
z: f16le(offset+((chunkLength*index)+(indey*8))+4)
});
//:skip 2
}
let normal = {
x: f16le(offset+(chunkLength*index)+((numberOfShapes * 8))),
y: f16le(offset+(chunkLength*index)+((numberOfShapes * 8))+2),
z: f16le(offset+(chunkLength*index)+((numberOfShapes * 8))+4)
};
//:skip 2
let uv = {
u: f16le(offset+(chunkLength*index)+((numberOfShapes * 8))+8),
v: f16le(offset+(chunkLength*index)+((numberOfShapes * 8))+10)
};
let color = u32_rgba8(u32le(offset+(chunkLength*index)+((numberOfShapes * 8))+12));
// keep original u32le color?
vertices.push({shapes, normal, uv, color});
}
offset = (20+(numberOfVertexes * chunkLength));
animationHeader = {id: u32le(offset), length: u32le(offset+4), speed: f32le(offset+8), "offset": u32le(offset+12), keyframes: u32le(offset+16)};
let animData = new Array();
// now we have to enumerate values, so now we introduce an offset value.
// format for a keyframe: sssskkkk[ffffvvvv] where [ffffvvvv] repeat based on the value that kkkk(eys) has.
// sssskkkk[ffffvvvv] is repeated based on animationHeader.keyframes value.
offset += 20;
for (let index = 0; index < animationHeader.keyframes; index++) {
let frameData = new Array();
let shapeId = u32le(offset);
let keys = u32le(offset+4);
offset += 8;
for (let indey = 0; indey < keys; indey++) {
frameData.push({frame: f32le(offset), value: f32le(offset+4)});
offset += 8;
}
animData.push({shapeId, keys, frameData});
}
let texture = null;
switch(textureFormat) {
case 'N': {
break;
}
case 'U': {
//where every 16-bit entry is a BGR5A1 color 0b[bbbbbgggggrrrrra]
texture = new Uint16Array(input.slice(offset, (offset+0x8000)));
//see convertBGR5A1toRGB5A1() for more info.
break;
}
case 'C': {
// compression format is RLE-based, where first u32 is size, and format is defined as:
/**
* u16 rleType;
* if (rleType >= 0xff00) {
* //do a raw copy
* let length = (0x10000 - rleType);
* byte data[length];
* } else {
* //repeat next byte rleType times
* data = new Uint16Array(rleType);
* for (let index = 0; index < rleType; index++) {
* data[index] = u16 repeater @ +4;
* }
* }
**/
//output of this will be another u16[0x4000] of the decompressed texture
//after that just parse output as-if it was uncompressed.
//see uncompressTexture() and convertBGR5A1toRGB5A1() for more info.
size = u32le(offset);
texture = {size, data: input.slice(offset+4, offset+(4+size))};
}
}
if(ICONJS_DEBUG){
console.debug({magic, numberOfShapes, textureType, textureFormat, numberOfVertexes, chunkLength, vertices, animationHeader, animData, texture});
}
return {numberOfShapes, vertices, textureFormat, texture, animData};
}
/**
* Read a 512-byte file descriptor that is used on Memory Cards or PSU files.
* @param {ArrayBuffer} input - File descriptor segment.
* @returns {Object} (user didn't write a description)
* @access protected
*/
function readEntryBlock(input) {
const {u32le, t64le} = new yellowDataReader(input);
//!pattern psu_file.hexpat
const permissions = u32le(0);
let type;
if (permissions>0xffff) {
throw `Not a EMS Memory Adapter (PSU) export file (was ${permissions}, expected less than ${0xffff})`;
}
if((permissions & 0b00100000)>=1){
type = "directory";
}
if((permissions & 0b00010000)>=1){
type = "file";
}
if((permissions & 0b0001100000000000)>=1){
throw `I don't parse portable applications or legacy save data. (${permissions} has bits 10 or 11 set)`;
}
const size = u32le(4);
const createdTime = t64le(8);
const sectorOffset = u32le(16);
const dirEntry = u32le(20);
const modifiedTime = t64le(24);
const specialSection = input.slice(0x20, 0x40);
const int_filename = input.slice(0x40, 512);
const filename = stringScrubber((new TextDecoder("utf-8")).decode(int_filename));
if(ICONJS_DEBUG){
console.debug({permissions, type, size, createdTime, sectorOffset, dirEntry, modifiedTime, specialSection, filename});
}
return {type, size, filename, createdTime, modifiedTime};
}
/**
* Read a EMS Memory Adapter export file (PSU format)
* @param {ArrayBuffer} input - PSU formatted file
* @returns {Object} (user didn't write a description)
* @public
*/
function readEmsPsuFile(input){
const header = readEntryBlock(input.slice(0,0x1ff));
if(header.size > 0x7f) {
throw `Directory is too large! (maximum size: ${0x7f}, was ${header.size})`
}
let fsOut = {length: header.size, rootDirectory: header.filename, timestamps: {created: header.createdTime, modified: header.modifiedTime}};
let output = new Object();
let offset = 512;
for (let index = 0; index < header.size; index++) {
fdesc = readEntryBlock(input.slice(offset, offset + 512));
switch(fdesc.type) {
case "directory": {
offset += 512;
output[fdesc.filename] = null;
break;
}
case "file": {
if(ICONJS_DEBUG){
console.debug(`PARSING | F: "${fdesc.filename}" O: ${offset} S: ${fdesc.size}`);
}
offset += 512;
const originalOffset = offset;
if((fdesc.size % 1024) > 0) {
offset += ((fdesc.size & 0b11111111110000000000) + 1024);
} else {
offset += fdesc.size;
// if we're already filling 1k blocks fully, why change the value?
}
output[fdesc.filename] = {
size: fdesc.size,
data: input.slice(originalOffset, originalOffset+fdesc.size)
};
break;
}
}
}
fsOut[header.filename] = output;
return fsOut;
}
/**
* Read a PS3 save export file (PSV format)
* @param {ArrayBuffer} input - PSV formatted file
* @returns {Object} (user didn't write a description)
* @public
*/
function readPsvFile(input){
const {u32le, t64le} = new yellowDataReader(input);
//!pattern psv_file.hexpat
const magic = u32le(0);
if (magic !== 0x50535600) {
throw `Not a PS3 export (PSV) file (was ${magic}, expected ${0x50535600})`;
}
//:skip 4
//:skip 20 // key seed, console ignores this
//:skip 20 // sha1 hmac digest, useful for verifying that this is, indeed, save data.
//:skip 8
const type1 = u32le(56);
const type2 = u32le(60);
if(type1 !== 0x2c && type2 !== 2) {
throw `Not parsing, this is not in the PS2 save export format (was ${type1}:${type2}, expected 44:2)`;
}
const displayedSize = u32le(64);
const ps2dOffset = u32le(68);
const ps2dSize = u32le(72); // don't know why this is included if its always 964
const nModelOffset = u32le(76);
const nModelSize = u32le(80);
const cModelOffset = u32le(84);
const cModelSize = u32le(88);
const dModelOffset = u32le(92);
const dModelSize = u32le(96);
const numberOfFiles = u32le(100); // in-case this library changes stance on other files
// file = {t64le created, t64le modified, u32 size, u32 permissions, byte[32] title}
// and if it's not the root directory, add another u32 for offset/location
const rootDirectoryData = input.slice(104, 162);
const timestamps = {created: t64le(104), modified: t64le(112)};
let offset = 162;
let fileData = new Array();
for (let index = 0; index < numberOfFiles; index++) {
fileData.push(input.slice(offset,offset+0x3c));
offset += 0x3c;
};
//then file data after this but we already have pointers to the files we care about
const icons = {
n: input.slice(nModelOffset, nModelOffset+nModelSize),
c: input.slice(cModelOffset, cModelOffset+cModelSize),
d: input.slice(dModelOffset, dModelOffset+dModelSize),
}
if (ICONJS_DEBUG) {
console.debug({magic, type1, type2, displayedSize, ps2dOffset, ps2dSize, nModelOffset, nModelSize, cModelOffset, cModelSize, dModelOffset, dModelSize, numberOfFiles, rootDirectoryData, fileData})
}
return {icons, "icon.sys": input.slice(ps2dOffset, ps2dOffset+ps2dSize), timestamps};
}
/**
* Read a SPS or XPS file descriptor.
* @param {ArrayBuffer} input - File descriptor segment.
* @returns {Object} (user didn't write a description)
* @access protected
*/
function readSxpsDescriptor(input) {
const {u32le, t64le} = new yellowDataReader(input);
//!pattern sps-xps_file.hexpat
//:skip 2 // ... it's the file descriptor block size (including the bytes themselves, so 250)
const int_filename = input.slice(2, 66);
const filename = stringScrubber((new TextDecoder("utf-8")).decode(int_filename));
const size = u32le(66);
const startSector = u32le(70);
const endSector = u32le(74);
const permissions = u32le(78); // the first two bytes are *swapped*. The comments that ensued were not kept.
let type;
if (permissions>0xffff) {
throw `Not a SharkPort (SPS) or X-Port (XPS) export file (was ${permissions}, expected less than ${0xffff})`;
}
if((permissions & 0b0010000000000000)>=1){
type = "directory";
}
if((permissions & 0b0001000000000000)>=1){
type = "file";
}
if((permissions & 0b00011000)>=1){
throw `I don't parse portable applications or legacy save data. (${permissions} has bits 4 or 5 set)`;
}
const timestamps = {created: t64le(82), modified: t64le(90)};
//:skip 4
//:skip 4 - u32 optional (98)
//:skip 8 - t64 optionalTime (102) // I don't know why this is here.
const int_asciiName = input.slice(114, 178);
const int_shiftjisName = input.slice(178, 242); // Because why parse a PS2D when you can hard-code it?
//:skip 8
if(ICONJS_DEBUG) {
console.debug({int_filename, size, startSector, endSector, permissions, type, timestamps, int_asciiName, int_shiftjisName});
}
return {type, size, filename, timestamps};
}
/**
* Read a SharkPort or X-Port export file (SPS or XPS format)
* @param {ArrayBuffer} input - SPS|XPS formatted file
* @returns {Object} (user didn't write a description)
* @public
*/
function readSharkXPortSxpsFile(input) {
const {u32le} = new yellowDataReader(input);
//!pattern sps-xps_file.hexpat
const identLength = u32le(0);
if(identLength !== 13) {
throw `Not a SharkPort (SPS) or X-Port (XPS) export file (was ${identLength}, expected 13)`;
}
let offset = 4;
const ident = input.slice(offset, offset+identLength);
if((new TextDecoder("utf-8")).decode(ident) !== "SharkPortSave") {
throw `Unrecognized file identification string. Expected "SharkPortSave".`
}
offset += (identLength + 4);
const titleLength = u32le(offset);
const title = input.slice(offset + 4, (offset + 4) + titleLength);
offset += (titleLength + 4);
const descriptionLength = u32le(offset);
const description = input.slice(offset + 4, (offset + 4) + descriptionLength);
offset += (descriptionLength + 4);
const description2Length = u32le(offset);
let description2;
if(description2Length !== 0) {
description2 = input.slice(offset + 4, (offset + 4) + description2Length);
offset += (description2Length + 4);
} else {
offset += 4;
}
const comments = {
"game": stringScrubber((new TextDecoder("utf-8")).decode(title)),
"name": stringScrubber((new TextDecoder("utf-8")).decode(description))
}
if(description2Length !== 0) {
comments.desc = stringScrubber((new TextDecoder("utf-8")).decode(description2));
}
const totalSize = u32le(offset);
offset += 4;
const header = readSxpsDescriptor(input.slice(offset, offset + 250));
offset += 250;
// alright now lets parse some actual data
let fsOut = {length: header.size, rootDirectory: header.filename, timestamps: header.timestamps, comments};
let output = new Object();
for (let index = 0; index < (header.size - 2); index++) {
fdesc = readSxpsDescriptor(input.slice(offset, offset + 250));
switch(fdesc.type) {
case "directory": {
offset += 250;
output[fdesc.filename] = null;
break;
}
case "file": {
if(ICONJS_DEBUG){
console.debug(`PARSING | F: "${fdesc.filename}" O: ${offset} S: ${fdesc.size}`);
}
offset += 250;
output[fdesc.filename] = {
size: fdesc.size,
data: input.slice(offset, offset+fdesc.size)
};
offset += fdesc.size;
break;
}
}
}
fsOut[header.filename] = output;
//:skip 4 // then here lies, at offset (the end of file), a u32 checksum.
return fsOut;
}
/**
* Define (module.)exports with all public functions.
* @exports icondumper2/icon
*/ // start c6js
exports = {
readers: {readIconFile, readPS2D, readEmsPsuFile, readPsvFile, readSharkXPortSxpsFile},
helpers: {uncompressTexture, convertBGR5A1toRGB5A1},
options: {setDebug, setStrictness},
version: ICONJS_VERSION
};
if(typeof module !== "undefined") {
module.exports = exports;
}
//end c6js
//start esm
/*export {
readIconFile, readPS2D, readEmsPsuFile, readPsvFile, readSharkXPortSxpsFile, uncompressTexture, convertBGR5A1toRGB5A1, setDebug, ICONJS_VERSION
};*/
//end esm