Tracking Issue for `const` `Iterator` impls and related `const fn`s

[fee1-dead]

Feature gate: #![feature(const_iter)]

This is a tracking issue for const Iterator impls and related const fns.

Note for interested contributors: please note that prior to an accepted const trait RFC, there is no need to constify everything related to Iterator. Please see #155816 for guidelines.

Public API

// core::iter
impl<I: ~const Iterator + ?Sized> const Iterator for &mut I;
impl<I: Iterator> const IntoIterator for I;
impl<B, I: ~const Iterator, F> const Iterator for Map<I, F>
where
    F: ~const FnMut(I::Item) -> B + ~const Drop;

// core::option
impl<T> Option<T> {
    pub const fn iter_mut(&mut self) -> IterMut<'_, T>;
}

impl<T> const IntoIterator for Option<T>;
impl<'a, T> const IntoIterator for &'a Option<T>;
impl<'a, T> const IntoIterator for &'a mut Option<T>;
impl<'a, A> const Iterator for Iter<'a, A>;
impl<'a, A> const Iterator for IterMut<'a, A>;
impl<A> const Iterator for IntoIter<A>;

Steps / History

• Implementation: implement const iterator using rustc_do_not_const_check #106541
• Final comment period (FCP)
• Stabilization PR

Unresolved Questions

• None yet.

[Expyron]

Does this issue also track const implementations of slice and str iterators and their related functions or would that be a separate tracking issue?

For example:

• core::slice::Iter and slice.iter()
• core::str::Chars and str.chars()

[fee1-dead]

@Expyron: those implementations have not been added yet. My plan is to land the smallest possible PR that simply makes it possible to work, and constify others in much smaller, finer grained PRs. Further work should use this tracking issue if related.

[TiagoCavalcante]

Any updates on this?

[fee1-dead]

@TiagoCavalcante: implementation has been stalled for more compiler related work to make this as easy to maintain as possible.

[ImUrX]

@fee1-dead could you update the linked PR to the new one?

[nazar-pc]

Slices and ranges are sorely missing for now