whatformat 0.2.0

detect fileformat by filename or while reading bytes

Released 2018-01-20.

To install, run:

haxelib install whatformat 0.2.0

See using Haxelib in Haxelib documentation for more information.

Maintainermaitag
Websitehttps://github.com/maitag/whatformat
Current version0.2.0
StatisticsInstalled 5 times
LicenseMIT
Tags

README.md

WhatFormat

pure haxe-library to detect fileformat by Magick Numbers) or by filename.

Installation

haxelib install whatformat

or use the latest developement version from github:

haxelib git whatformat https://github.com/maitag/whatformat.git

Documentation

Start with an instance of WhatFormat:

var wtf = new WhatFormat();

// or define what formats to check only:
// var wtf = new WhatFormat(['png', 'jpg', 'gif']); 

To check the Bytes in the header of a loaded File:

var bytes:Bytes = File.getBytes(filename);
if ( wtf.checkHeaderBytes(bytes).found ) trace('format: ${wtf.format} - ${wtf.description}');

For streams or special purpose you can check format while loading,
to easy stop loading if no fileformat can be detected by header:

var file:FileInput = File.read(filename);
var wtf:WhatFormat = new WhatFormat();
try {
	do {
		var byte:Int = file.readByte();
		// ... store it into Bytes or do whatever with
		
		wtf.checkNextByte(byte); // check next Byte in header (if formatcheck is in proceed)
	}
	while ( wtf.proceed || wtf.found); // stop reading if no format found
}
catch( ex:haxe.io.Eof ) {}

if (wtf.found) trace( 'format: ${wtf.format} - ${wtf.description}');
else trace("can't detect fileformat");

file.close();

You can also check the ending of a filename:

if (wtf.checkFilenameEnding(filename).found) {
	trace('format detected by filename ending:\n');
	trace('format: ${wtf.format} - ${wtf.description}');
	trace('subtype: ${wtf.format} - ${wtf.subtypeDescription}');
}

or combine both methods:

var wtf:WhatFormat = new WhatFormat();

// detect by filename
if (wtf.checkFilenameEnding(filename).found)
	trace('format detected by filename ending:\n', wtf.byName);

var input:Bytes;
var cache:BytesOutput = new BytesOutput();
var file:FileInput = File.read(filename);
var byte:Int;
try {
	do {
		byte = file.readByte();
		cache.writeByte(byte);
	}
	while ( wtf.checkNextByte(byte) || wtf.found); // do not stop if something found byName or byHeader
}
catch( ex:haxe.io.Eof ) {}
file.close();

// detected by header
if (wtf.byHeader.found) trace('format detected by parsing header:\n', wtf.byHeader);

// check again what filename-detection found
// if (wtf.byName.found) trace('format detected by filename - subtype:\n', wtf.byName.subtype);

// detected by byHeader OR byName
if (wtf.found) {
	input = cache.getBytes();
	trace('format of $filename found.');
	trace('filesize: ${input.length}');

	// if you not prefer the "header-first" autoselection method here,
	// set the second parameter in instantiation to false: wtf = new WhatFormat(null, false);
	trace('format: ${wtf.format} - ${wtf.description}');
	if (wtf.subtype != null) trace('subtype:${wtf.subtype} - ${wtf.subtypeDescription}');
}
else trace("can't detect fileformat");

For another check with the same instance of wtf,
in some cases you need to reset the detected results to defaults:

wtf.byHeader.reset();// resets wtf.byHeader values and get ready for another 'wtf.checkNextByte()'
wtf.byName.reset(); // resets wtf.byName values
wtf.reset();       // resets both

Supported Formats

Look here: src/formats/Magic.hx or test out Drag & Drop Sample in your Webbrowser --> here.

Feel free to commit new formats that needs a wtf-check;)

Todo

  • add more formats
  • check subtypes headers (container-formats, streams)