compress-integers

Compress a series of related integers.

Installation

npm i compress-integers -S

Usage

• on-line

IC = require('compress-integers');
IC.getCompressor({order: 2});
IC.write(10);
IC.write(10);
IC.getResult()

• off-line shortcut

buffer = IC.compress([1, 2, 3, 4, 5], {order:1});

• decompress

IC.decompress(buffer);

API

src

• getCompressor (opts = {})

  Get an order N compressor with options

  • param: opts { Object={} }

    options

  • option: RLE { boolean }

    use Run-Length Encoding

  • option: order { number=1 }

    order of the compressor (0, 1 or 2)

  • return: { Compressor }

    an order N compressor

• compress (arr, options = {})

  A shortcut for get a Compressor

  • param: arr { Array }

    array of integers

  • param: options { options={} }

    see getCompressor

  • return: { Buffer }

    compressed data

• decompress (buffer)

  Decompress a buffer to an array of integers

  • param: buffer { Buffer }

    buffer to compress

  • return: { Array }

    array of integers

• Compressor.write (int)

  write an integer

  • param: int { number }

    integer

• Compressor.getResult ()

  Get the result buffer

  • return: { Buffer }

    buffer

HOW

• Integers are encoded with zig-zag variable-length coding using binbone
• A series of integers are stored as their differences. Order is the number of turns of making difference.
• Run-length encoding is use to make it better.
• This algorithm is not suitable for random integers.

For example:

• [1, 2, 3, 4, 5]

  • order1: [1, 1, 1, 1, 1]
  • order2: [1, 1, 0, 0, 0] The second value is the base value.

• [1, 1, 1, 1, 1]

  • order1: [1, 0, 0, 0, 0]
  • order2: [1, 0, 0, 0, 0]

Comparison

npm run compare

See result

License

MIT@Jingchen Zhao