Replacing some loops with onEachLazy.

What is the difference? Why is onEachLazy better?

It's not generally better so to speak. I wrote onEachLzay because on chrome, when you use elements iterators (such as document.getElementsByTagName("a"), if you don't store the whole iterator into a vector, it has disastrous performance. I think the explanation was that the browser was rebuilding the iterator every X items you iterate through or if the DOM was updated. In both cases, since we don't care about the new elements, we can simply store the iterator into an array and avoid this issue.

Turn buildHelperPopup into a variable so it can be "replaced" once the function has been called once so it's not called again.

Didn't it do that already?

$ node
> function f(a) {}
undefined
> f
[Function: f]
> f = 1
1
> f
1

Why is the new code better?

Not better, just more clear (at least for me). I can remove this commit if you prefer?

Not better, just more clear (at least for me). I can remove this commit if you prefer?

Yes, I think I prefer the old code. Thanks.

r=me with that done

@bors: r=jyn514 rollup

📌 Commit 8b6304e has been approved by jyn514

⌛ Testing commit 8b6304e with merge 0cd459f...

☀️ Test successful - checks-actions
Approved by: jyn514
Pushing 0cd459f to master...