A set of NodeJS streams aimed on chunking data
JavaScript
Permalink
Failed to load latest commit information.
lib SizeChunker: No more unnecessary empty chunks emitted Jun 14, 2015
test SizeChunker: No more unnecessary empty chunks emitted Jun 14, 2015
.gitignore Initial commit Oct 6, 2013
.travis.yml Build: Enable building on 0.12 and latest io.js Jun 14, 2015
LICENSE Initial commit Oct 6, 2013
README.md
index.js FilterStream Aug 13, 2014
package.json SizeChunker: No more unnecessary empty chunks emitted Jun 14, 2015

README.md

node-chunking-streams

Build Status Coverage Status NPM version Dependency Status

A set of Node.js streams to process data in chunks

  1. LineCounter
  2. SeparatorChunker
  3. SizeChunker
  4. GzipChunker
  5. S3MultipartUploader

To intstall with NPM

npm install chunking-streams

To use with Node.JS

var chunkingStreams = require('chunking-streams');

var LineCounter = chunkingStreams.LineCounter;
var SeparatorChunker = chunkingStreams.SeparatorChunker;
var SizeChunker = chunkingStreams.SizeChunker;

// ad so on...

LineCounter

Simple TransformStream which counts lines (\n is a separator) and emit data chunks contains exactly specified number of them.

Configuration

new LineCounter({
    numLines: 1,        // number of lines in a single output chunk. 1 is default
    flushTail: false    // on stream end, flush or not remaining buffer. false is default
});

SeparatorChunker

Split incoming data into chunks based on specified separator. After each separator found data chunk is emitted. By default separator sequence is set to \n, so it is equals to LineCounter with numLines: 1

Configuration

new SeparatorChunker({
    separator: '\n', // separator sequence
    flushTail: false // on stream end, flush or not remaining buffer. false is default
});

SizeChunker

Split streams into chunks having exactly specified number of bytes. It is object mode stream! Each data chunk is an object with the following fields:

  • id: number of chunk (starts from 0)
  • data: Buffer with data

SizeChunker has 2 additional events:

  • chunkStart: emitted on each chunk start.
  • chunkEnd: emitted on each chunk finish.

Both event handlers must accept two arguments:

  • id: number of chunk
  • done: callback function, must be called then processing is completed

If some tail data is not fit fully into specified chunk size, it can be emitted at the end if flushTail flag is set.

Configuration

new SizeChunker({
    chunkSize: 1024 // must be a number greater than zero.
    flushTail: true // flush or not remainder of an incoming stream. Defaults to false
});

Example

var input = fs.createReadStream('./input'),
    chunker = new SizeChunker({
        chunkSize: 1024
    }),
    output;

chunker.on('chunkStart', function(id, done) {
    output = fs.createWriteStream('./output-' + id);
    done();
});

chunker.on('chunkEnd', function(id, done) {
    output.end();
    done();
});

chunker.on('data', function(chunk) {
    output.write(chunk.data);
});

input.pipe(chunker);

S3MultipartUploader

INCOMPLETE