Run length encoder/decoder written in C.
This compressor uses positive- and negative-run RLE to lossleslly encode files.
- Positive counts represent the run length of repeated characters (i.e.
aaaaencodes to4a). - Negative counts represent the run length of non-repeating characters (i.e.
abcdencodes to-4abcd). These negative counts minimize the number of bytes needed to encode a long non-repeating run (only 1 byte needed per 127 characters instead of the 1 byte needed for each character if only positive runs were used).
All counts are represented in encoded files as signed, 8-bit integers (-128 to 127).
Note: Compressed files will be saved with a .rle extention. foo.txt compresses to foo.txt.rle. Decompressed files will be saved with the original filename. If a .rle extention is found, it will be removed, potentially overwriting a file with the same name. All files will be written to the same folder as the source file.
Information entropy and optimal size after compression is calculated using Shannon's Theorem.
make./RLE MODE FILENAME: Mode is 1 (for compression) or 2 (for decompression)
- Tested on Mac
- Tested on Linux
- Tested on Windows
- Dynamic realloc for decoding (instead of precomputing worst case)
- #define and use constants