aPLib is a compression library based on the algorithm used in aPACK (my 16-bit executable packer). aPLib is an easy-to-use alternative to many of the heavy-weight compression libraries available.
The compression ratios achieved by aPLib combined with the speed and tiny footprint of the decompressors (as low as 169 bytes!) makes it the ideal choice for many products.
Since the first public release in 1998, aPLib has been one of the top pure LZ-based compression libraries available. It is used in a wide range of products including executable compression and protection software, archivers, games, embedded systems, and handheld devices.
The aPLib package includes pre-compiled libraries in a number of formats (COFF, ELF, OMF).
No standard library functions are used, so the libraries can work with most x86/x64 compilers, as long as the name decoration and calling conventions match.
The ELF folders contain a version of aPLib which uses PIC to allow it to be linked into shared libraries on linux.
All compression and decompression functions are thread-safe.
For C/C++ you simply include aplib.h and link with the appropriate library for your compiler. If you only need to decompress data, or if you modify the decompression code, you can compile and link with one of the decompression implementations in the src folder.
For other languages you can either check if there is a useable example, or use the DLL version. Most linkers allow calling C functions in an external library, so usually there is a way to use one of the libraries.
aPLib performs memory-to-memory compression and decompression, so getting data into an input buffer and allocating an output buffer is your responsibility.
All functions return APLIB_ERROR (which is -1) if an error occurs.
Attempting to compress incompressible data can lead to expansion. You can get the maximum possible coded size by passing the size of the input to the function aP_max_packed_size().
When calling aP_pack() you have to supply a work buffer. You can get the required size of this buffer by passing the size of the input to the function aP_workmem_size() (in the current version this function always returns 640k).
If you do not have a callback for aP_pack(), you can pass NULL (i.e. 0) instead. The callback functionality allows your program to keep track of the compression progress, and if required to stop the compression.
aP_depack(), aP_depack_asm(), and aP_depack_asm_fast() assume that they are given valid compressed data – if not they will most likely crash. This is to ensure that the basic decompression code is as small, fast and easy to understand as possible. You can use aP_depack_safe() or aP_depack_asm_safe() if you need to catch decompression errors. Also the safe wrapper functions provide a nice interface that helps prevent potential crashes.
Safe Wrapper Functions
Starting with aPLib v0.34, there are additional functions included which provide a better way of handling the compressed data in the example, and also serve as an example of how to add functionality through function wrappers.
The aPsafe_pack() and aPsafe_depack() functions are wrappers for their regular aP_ counterparts, which add a header to the compressed data. This header includes a tag, information about the compressed and decompressed size of the data, and CRC32 values for the compressed and decompressed data.
The example folder contains a simple command line packer that uses aPLib to compress and decompress data. The aPsafe_ functions are used in this example, because they provide extra functionality like retrieving the original size of compressed data.