mcollina/split2

Split2(matcher, mapper, options)

Break up a stream and reassemble it so that each line is a chunk. split2 is inspired by @dominictarr split module, and it is totally API compatible with it. However, it is based on Node.js core Transform.

matcher may be a String, or a RegExp. Example, read every line in a file ...

  fs.createReadStream(file)
    .pipe(split2())
    .on('data', function (line) {
      //each chunk now is a separate line!
    })

split takes the same arguments as string.split except it defaults to '/\r?\n/', and the optional limit paremeter is ignored. String#split

split takes an optional options object on it's third argument, which is directly passed as a Transform option.

Additionally, the .maxLength and .skipOverflow options are implemented, which set limits on the internal buffer size and the stream's behavior when the limit is exceeded. There is no limit unless maxLength is set. When the internal buffer size exceeds maxLength, the stream emits an error by default. You may also set skipOverflow to true to suppress the error and instead skip past any lines that cause the internal buffer to exceed maxLength.

Calling .destroy will make the stream emit close. Use this to perform cleanup logic

var splitFile = function(filename) {
  var file = fs.createReadStream(filename)

  return file
    .pipe(split2())
    .on('close', function() {
      // destroy the file stream in case the split stream was destroyed
      file.destroy()
    })
}

var stream = splitFile('my-file.txt')

stream.destroy() // will destroy the input file stream

NDJ - Newline Delimited Json

split2 accepts a function which transforms each line.

fs.createReadStream(file)
  .pipe(split2(JSON.parse))
  .on('data', function (obj) {
    //each chunk now is a js object
  })
  .on("error", function(error) {
    //handling parsing errors
  })

However, in @dominictarr split the mapper is wrapped in a try-catch, while here it is not: if your parsing logic can throw, wrap it yourself. Otherwise, you can also use the stream error handling when mapper function throw.

Benchmark

The benchmark reads package.json and splits it into lines 10,000 times with split2, binary-split, and Node.js readline.createInterface. Run it with:

npm run benchmark -- --no-color

Example results on Node.js 24.18.0, Linux x64, with an Intel Core i7-7700:

benchSplit*10000: 714.092ms
benchBinarySplit*10000: 866.643ms
benchCreateInterface*10000: 869.476ms
benchSplit*10000: 683.199ms
benchBinarySplit*10000: 772.226ms
benchCreateInterface*10000: 739.976ms

Results will vary by machine and Node.js version.

License

Copyright (c) 2014-2021, Matteo Collina [email protected]

Permission to use, copy, modify, and/or distribute this software for any purpose with or without fee is hereby granted, provided that the above copyright notice and this permission notice appear in all copies.

THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.