A fast, high performance, lossless compression system.

Code documentation is here: https://jaredmales.github.io/xrif-docs/modules.html

The only dependencies are the check unit testing framework and subunit streaming system.

On CentOS:

$ sudo yum install check-devel
$ sudo yum install subunit-devel

On Ubuntu:

$ sudo apt-get install check

For a standard posix system, you should only need to do these steps:

$ mkdir build 
$ cd build 
$ cmake ..
$ make
$ make test
$ sudo make install
$ sudo ldconfig

The below code shows the steps needed to initialize an xrif handle and use it to compress a cube of images.

int16_t * source_data;
// ... allocate and populate source_data with `planes` images of size `rows X cols`

xrif_t xrif;
xrif_error_t rv;
rv = xrif_new(&xrif);

rv = xrif_set_size(xrif, rows, cols, 1, planes, XRIF_TYPECODE_INT16); // For a cube with `planes` images of size `rows X cols` of type `int16_t`

rv = xrif_allocate(xrif);

memcpy( xrif->raw, source_data, rows*cols*planes*sizeof(int16_t));

rv = xrif_encode(xrif);

rv = xrif_decode(xrif);

memcpy(source_data, xrif->raw, rows*cols*planes*sizeof(int16_t));

The library provides ways to use source_data directly (avoiding the memcpy), and control over which algorithms are used at each step of the compression.

Note that only int16 monochrome images are fully supported at this time.

The current version is 0

The size of the data is specified by width X height X depth X xrif_typesize(typecode) X frames

Difference method can be:

Reorder method can be:

Compression method can be:

If Compression method is LZ4 then bytes 40-41 are uint16_t containing the lz4_acceleration parameter.

The code documentation is here: https://jaredmales.github.io/xrif/