Arithmetic coding is the loss-less compression algorithm which works the best by a theoretical bound.
When encrypting the current symbol, we have to predict how distribution of the real data will change with time, the better we do that - the better the compression will be.
Me
--input <file> = input file, default = `input.txt`.
--output <file> = output file, default = `output.txt`.
--mode {c | d} = 'c' ~ compression, 'd' ~ decompression, default = 'd', .
--method {ari | ppm} = 'ari' - arithmetic coding, 'ppm' - prediction by partial matching, default = 'ari'.
# Unix
archiver --input the_financier.txt --output the_financier.cmp --mode c --method ppm
# Windows
archiver --input the_financier.txt --output the_financier.cmp --mode c --method ppm