Example code for processing Gzipped file in Node.js using node-bufferstream

process_ntriple_gz.coffee

fs = require('fs')
zlib = require('zlib')

# see: https://github.com/dodo/node-bufferstream
BufferStream = require('bufferstream')

input = fs.createReadStream inputPath,
  flags: 'r'
  mode: 0666

gunzip = zlib.createGunzip()

# assume output is lines of UTF-8 (i.e. RDF N-Triples)
output = new BufferStream
  encoding: 'utf8'
  size: 'none'
output.split '\r\n', '\r', '\n'
output.on 'split', (line, token) ->
  # line is actually always Buffer despite encoding option
  line = line.toString() if typeof line isnt 'string'
  console.log line

input.pipe(gunzip).pipe(output)