Add ParallelIterator::collect_vec_list

[coolreader18]

I wasn't exactly sure where the collect_vec_list implementation should go, since the necessary internals are in extend but it's logically a collect function. Also, I feel like the documentation could be better, I'd appreciate feedback on that.

Resolves #1127

[coolreader18]

Wasn't sure if I should re-use pi.collect::<Vec<_>>() here or not; theoretically that'd have an unnecessary second check of opt_len() but that would get optimized in almost all cases.

[cuviper]

The way you've done it is fine, especially since that internal method is already shared pub(super).

The other thing I noticed is that the unindexed case avoids pushing empty vecs to the list, but here it will always create that one node. I'm not sure that matters though -- it's potentially more meaningful when there could end up being many empty list nodes, instead of just this one in a degenerate case.

[martin-g]

Suggested change

	let mut v = Vec::new();
	let mut v = Vec::with_capacity(len);

[cuviper]

This suggested change is fine, but I think it won't really matter since special_extend -> collect_with_consumer will call reserve anyway. Allocate now or later, but we're still only doing it once.

[coolreader18]

Right, that was my thinking, and the ParallelExtend impl doesn't reserve here either.

[cuviper]

I wasn't exactly sure where the collect_vec_list implementation should go, since the necessary internals are in extend but it's logically a collect function.

I think that bit of code is short enough that it could just be implemented directly in the trait method.

[coolreader18]

Something I was thinking of including in the docs but wasn't sure about - mentioning the pattern of calling .into_iter().flatten() on the linked list? Like, .collect_vec_list().into_iter().flatten() is now a very straightforward/efficient way of turning a parallel iterator into a serial one, though I'm unsure if that should be encouraged or where in the docs that would go.

[coolreader18]

Also, added a check for length of 0 returning an empty LL, though yeah I'm not sure if that's necessary.

[cuviper]

Looks good, thanks!

[coolreader18]

Would it be possible to cut a release with this?

[cuviper]

This is now published in rayon 1.9.0.

[coolreader18]

Thank you so much!