For anyone wanting to try this at home, the contraption uses a Huffman binary code.

To make the code a canonical Huffman code, the codes are renumbered.

Huffman threaded code consists of lists of Huffman codes.

A Huffman code is a variable length bit string used to identify a unique item.

To reduce the size of a file by a complicated scheme that produces bit configurations completely unrelated to the original data, such as by a Huffman code.

Universal codes are also useful when Huffman codes are inconvenient.

Parallels with Huffman codes and probabilistic binary search trees are drawn.

A method was later found to design a Huffman code in linear time if input probabilities (also known as weights) are sorted.

Note that, in general, a Huffman code need not be unique, but it is always one of the codes minimizing .

The resulting raw data is compressed using a modified Huffman code optimized for written text, achieving average compression factors of around 20.