hklindworth / format

Various files formats support for Haxe

The format library contains support for different file-formats for the Haxe programming language.

Formats

Currently supported formats are :

• ABC (Flash AS3 bytecode format)
• AGAL (Adobe Shader Assembler for Stage3D - writer only atm)
• AMF (Flash serialized object)
• AS1 (Adobe ActionScript1-2 bytecode in SWF - reader only atm)
• BMP (Bitmap Image format)
• FLV (Flash Video)
• GZ (compressed file)
• JPG (Image file format - writer only atm)
• MP3 (compressed audio)
• NEKO (NekoVM bytecode)
• PBJ (PixelBender Binary file)
• PDF (only generic file structure and partial decryption)
• PEX (particle effect format)
• PNG (image file format)
• SWF (Flash file format)
• TAR (Archive)
• TGZ (TAR+GZ Archive)
• WAV (Raw sound)
• ZIP (Compressed Archive)
• GIF (Image file format - writer don't have tools)
• TGA (TARGA Image file format; Reader/Writer does not support developer data chunk; Writer does not support RLE encoding)

Documentation

Automatically generated API documentation is here: http://haxefoundation.github.io/format/

Installation

Available on haxelib, simply run the following command : haxelib install format. To use the library, simply add -lib format to your commandline parameters.

Package Structure

Each format lies in its own package, for example format.pdf contains classes for PDF.

The format.tools package contain some tools that might be shared by several formats but don't belong to a specific one.

Each format must provide the following files :

• one Data.hx file that contain only data structures / enums used by the format. If there is really a lot, they can be separated into several files, but it's often my easy for the end user to only have to do one single import format.xxx.Data to access to all the defined types.
• one Reader.hx class which enable to read the file format from an haxe.io.Input
• one Writer.hx class which enable to write the file format to an haxe.io.Output
• some other classes that might be necessary for manipulating the data structures

It's important in particular that the data structures storing the decoded information are separated from the actual classes manipulating it. This enable full access to all the file format infos and the ability to easily write libraries that manipulate the format, even if later the Reader implementation is changed for example.

Contributing

We're accepting contributions if they are following the package structure rules (see above), please send them as Pull Requests.