Fix ArrayBufferView, ArrayOps.ArrayIterator and View.Patched

[NthPortal]

Ensure ArrayBufferView is consistent with its buffer.

Simplify ArrayBuffer#insertAll code.

Fix unreported bug in ArrayOps.ArrayIterator where
index in array can overflow to negative, and in such cases
the iterator reports hasNext incorrectly as true.

Improve code documentation and readability of
Iterator.patch implementation, which was bumped
into while investigating the bug.

Discovered while writing tests for bug in View.Patched
(see below).

Fix unreported bug in View.Patched where iterator
is incorrect due to the patch only being iterable once,
and already having been exhausted.

Discovered while attempting to optimise ArrayBuffer alongside
fixing ArrayBufferView (eventually scrapped).

Fixes scala/bug#12284

[julienrf]

Thank you so much @NthPortal! (for your patience also…!)

[lrytz]

LGTM, again leaving to @SethTisue

[SethTisue]

@NthPortal hard to be totally sure, but afaict the ArrayBufferBenchmark I added over in #9258 should suffice for benchmarking this one as well

[NthPortal]

@SethTisue yes.

I'll have to review the changes, as it's been a really long time since I've looked at them, but I'm not even sure there's anything to benchmark

edit: we need to benchmark addAll. I don't think the current ArrayBufferBenchmark is sufficient, as we need to benchmark changes to the performance of addAll(IterableOnce), addAll(Iterable) and addAll(ArrayBuffer). I believe the benchmark currently only runs addAll(ArrayBuffer)

[SethTisue]

2nd attempt to attract reviewer attention from @scala/collections crew and/or @julienrf

[julienrf]

Hey, sorry I can’t review the PR soon.

[johnynek]

maybe I suffer from anxiety, but given this is in the standard library, and given that we already have found bugs, I think we need more extensive tests. I've suggested two scalachecks to add.

[SethTisue]

@NthPortal what's next here? are you interested in doing the ScalaCheck thing @johnynek suggested?

[NthPortal]

@SethTisue i've been and continue to be sick, but i'll try to add the scalacheck stuff today or tomorrow. if i don't get to it, and @johnynek has perms, they can feel free to add a commit to this branch with it

[NthPortal]

oops I've bumped into ANOTHER unreported bug (by way of testing the fix for the first unreported bug too, not just separately as part of working on the main bug)

[NthPortal]

this is done as far as I'm concerned

[NthPortal]

@johnynek how do you feel about the tests now?