This repository contains a compression algorithm written by me (Michael Grigoryan). The algorithm is only able to compress and decompress text files and is not guaranteed to work with other file types.

The algorithm works best with repetitive texts.

Suppose you have this repetitive text:

Nory was a Catholic because her mother was a Catholic,
and Nory’s mother was a Catholic because her father
was a Catholic, and her father was a Catholic because
his mother was a Catholic, or had been.

Taken from https://thejohnfox.com/2021/08/17-fantastic-repetition-examples-in-literature/

As you can see the text is very repetitive. The compression algorithm will loop through all the lines and the words in them.

If a word is not present in the shared index and is used multiple times throughout the text body, the algorithm will append that word to the shared index. If the word is only used once, then it is added directly to the compressed file, without being added to the shared index.

After the compression the file will produce a compressed.mca file which will have the following content:

["Nory","was","a","Catholic","because","her","mother","Catholic,","and","father","or"]
0 1 2 3 4 5 6 1 2 7
8 Nory’s 6 1 2 3 4 5 9
1 2 7 8 5 9 1 2 3 4
his 6 1 2 7 10 had been.

The first line of the output contains the shared index. If you attempt to decompress a file without this "header" the program will throw a corruption error.

You can start the program by running:

cargo run --release

after which you should get a prompt, asking you to choose an action:

select an option: (c)ompress/(d)ecompress:

Both, compression and decompression are supported. Compressing a file will create a file named compressed.mca and decompression will output a file named decompressed.txt.