uhbc

Quotes from the program documentation:

UHBC is a blocksorting compressor optimized for high compression ratios. The current version is still in an experimental state, but because the results are close to the world's best, I decided to release this program as public BETA version. It is still under development, so you use this program at your own risk. UHBC doesn't use any data filters or transformations before BWT.

UHBC is a file-to-file-compressor, not an archiver. So it doesn't store name, time/date or attributes of the original file. There is no check for existing files and there is also no file verification after decompression.

Switches:

• -b specify maximum block size in bytes, kilobytes (k) or megabytes (m), e.g. -b16m or -b900k or -b4096 default: -b5m, maximum: -b128m, size is rounded to 1kb
• -m choose second stage algorithms, default: -m2
  • -m0 RLE + direct entropy coding
  • -m1 RLE + MTF + entropy coding
  • -m2 RLE + WFC + entropy coding
  • -m3 RLE + DWFC + entropy coding
• -s process input with second stage algorithms only (don't perform BWT); input should already be processed by BWT but it will work for any data
• -d disable adaptive switching between MTF/WFC and direct coding; it is faster and maybe useful for comparisons
• -c choose contexts for BWT sorting, default: auto-selection
  • -cf use following contexts
  • -cp use preceding contexts

  Memory requirements:

  • compression 8*blocksize + 256kb
  • decompression 5*blocksize

The compressor is featured in the BWT comparison.

(Have info that should be added here? E-mail.)