split one iterable into two based on a test function


201.0.14 years ago4 years agoMinified + gzip package size for @iterables/sieve in KB



Sometimes you want to iterate over a set once and split the set into two separate usable piles.

This does that!

const sieve = require('@iterables/sieve')

const [yes, no] = sieve([1, 0, 1], (xs, idx, all) => {
  return Boolean(xs)

console.log([...yes]) // [1, 1]
console.log([...no]) // [0]

// works with any iterable:
const [yes, no] = sieve(function * () {
  yield 0
  yield 1
  yield 2
}(), xs => Boolean(xs))

console.log([...yes]) // [1, 2]
console.log([...no]) // [0]


$ npm install --save @iterables/sieve


sieve(iterable:Iterator<T>, test:Function) -> [Iterator<T>, Iterator<T>]

  • iterable: any iterable (generator instance, Set, Map, Array, etc.)
  • test: a function taking xs, idx, and all, returning Boolean
    • xs: an item from iterable.
    • idx: the index within the array.
    • all: the original iterable

Returns a two-element array of iterators. The first element, yes, is an iterable of all items from the original iterable that pass the test function. The second element, no, contains all the failing items.

Note: this function will buffer the skipped items in the second iterable to be realized. That is, if you evaluate the yes iterable all at once first, all failing elements will be buffered in no until it is iterated. Usually realizing the array is the goal, so keeping the elements in memory shouldn't be an issue.

If memory is an issue, consider taking one element at a time from yes and no, which will buffer at most N items, where N is the largest run of all-passing or all-failing items.



If you find any bugs or have a feature request, please open an issue on github!

The npm package download data comes from npm's download counts api and package details come from npms.io.