Question

I’m filtering a given Array using its native .filter() method:

var a = [1,2,3,4,5];

var b = a.filter(function(v) {
    return v > 2;
});

That creates the new Array ([3,4,5]). Now, I also want to have the filtered values in another array. What is my best option here? Should I

• push the removed values into a new array within that same filter method ?
• write an invert function and apply it after filtering ?

To go with the first option, it might end up this:

var b = a.filter(function(v) {
    return v > 2 || !c.push(v);
});

My problem with that solution is, that it kinda mixes two different things and maybe very confusing for anybody who reads the code in the future. As an alternative, I could call something like

c = invert(a,b);

function invert(source, compare) {
    return source.filter(filterPositives);

    function filterPositives(v) {
        return compare.indexOf(v) === -1;
    };
}

Is that effective? Or can I do better ?

Any other (more elegant) ideas how to solve this problem ?

Answer 1

I don’t think going through the source array twice is an elegant solution. But then again, adding side-effects to your own filter is not great either.

I’d solve the problem by writing a filterSplit function, something like this pseudocode:

function filterSplit(
      Array source, Array positives, Array negatives, Function filterCb) 
{
  source.forEach(function(el) {
    if (filterCb(el)) {
      positives.push(el); 
    }
    else {
      negatives.push(el);
    }
  }
}

… or, if you’d prefer arrays in your return …

function anotherFilterSplit(Array source, Function filterCb) {
  var positives = [], negatives = [];
  // ... the same check and push as above ...
  return [positives, negatives];
}