I wonder if we should have this return a record { before :: Array a, after :: Array a } rather than a Tuple, since we're already doing something similar with partition, and the splitAt for strings already returns a record: https://pursuit.purescript.org/packages/purescript-strings/4.0.1/docs/Data.String.CodePoints#v:splitAt

I'd agree with that. It also removes the need to document which array is the before/after part (though that itself is likely obvious).

splitAt now returns { before :: Array a, after :: Array a } instead of a Tuple.

Funnily enough, I had initially considered doing this, but was hung up on the before terminology since I was thinking of it in terms of take n elements, which is inclusive.

It wasn't until just now where I realized that, conceptually, taking n elements from an array (inclusive) is the same as taking elements up to that index (exclusive).

It wasn't until just now where I realized that, conceptually, taking n elements from an array (inclusive) is the same as taking elements up to that index (exclusive).

I see it more like this:

let { before, after} = splitAt i array
in 
  (before == take i array) && (after == drop i array)