CODEWITHSUNDEEP
javascriptrubynode.jszlib
Brandon - Free Palestine
Brandon - Free Palestine

Reputation: 16666

Node.js zlib.deflate output too long

I'm trying to convert a Ruby script I wrote into Node.js, but my zlib compression output is way to big (1580 bytes instead of the 62 ruby gives me). What I'm compressing is very uniform, 10,000 bytes of data which can either be 65 or 122.

Ruby:

rimg = []
(1..100).each do |i|
  (1..100).each do |j|
    rimg << ((j > 20 && j < 30) || (j > 70 && j < 80) ? 65 : 122)
  end
end

zd = Zlib::Deflate.new(Zlib::DEFAULT_COMPRESSION, Zlib::MAX_WBITS)
ximg = zd.deflate(rimg.pack('c*'), Zlib::FINISH)
zd.close
puts "bytesize #{ximg.bytesize}"

Bytesize is 62 bytes. Here is my JS:

  var rimg = new Buffer(100*100);
  rimg.fill(0);

  for (var i = 1; i <= 100; i++) {
    for (var j = 1; j <= 100; j++) {
      rimg.writeInt8((((j > 20 && j < 30) || (j > 70 && j < 80) ? 65 : 122)), i*j-1);
    }
  }

  Zlib.deflate(rimg, { windowBits: 15 }, function(err, result) {
    console.log('-------')
    console.log(result.length);
  });

This outputs 1580. Node version is 5.1.1. What is the difference?

The header bytes for both are "78 9C" so they should be using the same compression

Upvotes: 0

Views: 147

Answers (1)

Mark Adler
Mark Adler

Reputation: 112482

It seems that you did not check your node.js buffer after your for loops to see if it really was what you thought it was. Your offset is i*j-1? Maybe you meant i*100 + j - 101? Or better yet, to avoid needless computations, just set k = 0 before the for loops and make the offset k++.

Upvotes: 1

Related Questions