Tracking issue for VecDeque Extras

[Gankra]

This covers the deque_extras blanket feature, which should probably be sharded more in the future. It covers:

• truncate
• resize
• as_slices
• as_mut_slices
• swap_back_remove
• swap_front_remove

Some notes:

• truncate and resize are strange because they aren't directional (they just truncate off the back). Need work.
• as_slices and as_mut_slices seem good to go. A bit niche, but so is VecDeque. Maybe move the mut to the end. RFC'd and everything.
• swap_back_remove and swap_front_remove also seem good to go. Maybe move the directional specifier to the end. RFC'd and everything.

[eefriedman]

shrink_to_fit and insert are now also part of deque_extras.

[sfackler]

I think we can pretty easily stabilize as_slices and as_mut_slices. Not sure about the others.

[alexcrichton]

This issue is now entering its cycle-long FCP for stabilization in 1.5

Specifically the as_slices and as_mut_slices along swap_*_remove methods are being considered for stabilization. We are also considering renaming the swapping methods.

[llogiq]

@alexcrichton How should the swapping methods be named?

[alexcrichton]

@llogiq as mentioned in the issue description, the other candidate names would be swap_remove_back and swap_remove_front

[alexcrichton]

The libs team discussed this during triage today and the decision was to stabilize:

• insert
• shrink_to_fit
• as_slices
• as_mut_slices
• swap_remove_front (renamed from swap_front_remove)
• swap_remove_back (renamed from swap_back_remove)

[Gankra]

Oh, we forgot to discuss as_mut_slices vs as_slices_mut.

I think we should move muts to the back as much as possible, even if in this case we're making "mutable slices", it's still the _mut variant of as_slices.

[alexcrichton]

Hm I'm not sure I'd follow the same reasoning, in method names types like &mut [T] and *mut T are referred to as mut_slice and mut_ptr. We've always known about this with regards to the _mut-at-the-end convention, so I would feel that these methods are correctly named.

[Gankra]

Yeah I've been a bit ambivalent about that naming scheme, but I guess it's
a bit too late for that!

On Thu, Oct 22, 2015 at 11:18 AM, Alex Crichton notifications@github.com
wrote:

Hm I'm not sure I'd follow the same reasoning, in method names types like &mut
[T] and *mut T are referred to as mut_slice and mut_ptr. We've always
known about this with regards to the _mut-at-the-end convention, so I
would feel that these methods are correctly named.

—
Reply to this email directly or view it on GitHub
#27788 (comment).