Tennix Archive Format (Version 1.0)

This space is reserved for the specification of the Tennix Archive Format. The current version of this archive format is 1.0.

• The 32-bit integers are saved in network byte order
• Tennix archive files always start with thpinfo.com/2009/tennix/afmt
• The item header data is xor'ed with the key byte from the header
• The item payload data is xor'ed with the key byte from the item header data
• Limits: 86 character filenames, 255 files per archive
• Version info is stored as two unsigned bytes (major=byte 31, minor=byte 32)

Usage

This file format is used by Tennix to put all resources (graphics, sounds, etc..) into a single container file that can be shipped alongside the executable. In case the code changes, but the data stays the same, only the (relatively small) executable file has to be exchanged. A bit of obfuscation is added to make the data appear as a single blob and prevent normal applications from interpreting this file in a wrong way. The process for opening and writing Tennix Archive Format-encoded files is described in the reference implementation below.

Reference implementation

The reference implementation is available in the Tennix Git repository and source tarballs of versions that include support for the Tennix Archive Format. It is also placed here in case you need it (be aware that the version in the Tennix repository might be more up-to-date than this local copy here):

• archive.c
• archive.h

thp.io | contact information