util: add JoinMap::keys

[andreastedile]

In a library I am developing, I need to know what tasks are contained in a JoinMap.

[hawkw]

it's definitely much simpler to return an impl Iterator here. However, I will note that the hash_map::Keys iterator that this wraps also implements some other traits: in addition to being an Iterator, it also implements Clone, ExactSizeIterator, and FusedIterator. It might be better if we returned an iterator that also implements those traits.

In particular, ExactSizeIterator is used by the standard library to implement optimizations when collect()ing an iterator into particular collections, like Vecs and HashMaps, since the size of the iterator can be reserved up front by with_capacity, instead of allocating multiple times while collecting.

Therefore, we may want to change this to return a named Keys iterator, like this:

use std::iterator::{ExactSizeIterator, FusedIterator};

pub struct Keys<'map, K, S = RandomState> {
    inner: hashbrown::hash_map::Keys<'map, Key<K>, AbortHandle, S>>,
}

impl<'map, K, S> Iterator for Keys<'map, K, S> {
    type Item = &'map K;

    fn next(&mut self) -> Option<Self::Item> {
         self.inner.next.map(|key| &key.key)
    }

    fn size_hint(&self) -> (usize, Option<usize>) {
        self.inner.size_hint()    
    }
}

impl<'map, K, S> ExactSizeIterator for Keys<'map, K, S> {
    fn len(&self) -> usize {
         self.inner.len()
    }

    fn is_empty(&self) -> bool {
         self.inner.is_empty()
    }
}

impl<'map, K, S> FusedIterator for Keys<'map, K, S> {}

Or, alternatively, I think we could write

Suggested change

	pub fn keys<Q: ?Sized>(&self) -> impl Iterator<Item = &K>
	pub fn keys<Q: ?Sized>(&self) -> impl Iterator<Item = &K> + ExactSizeIterator + FusedIterator

but I think the named type might be a nicer API surface for users?

[Darksonn]

Thank you for the PR. I made a few changes according to Eliza's review.

@andreastedile Since it will be attributed to you, please let me know whether you are okay with me merging the current contents of this PR.

[andreastedile]

@Darksonn Yes, I am okay with it: I agree with the suggestions and modifications. Thank you both for dedicating time to this!

[Darksonn]

I'll make a release containing this once #6033 is merged.