Tracking issue for Vec::resize

[alexcrichton]

This is a tracking issue for the unstable vec_resize feature in the standard library. To the best of my knowledge there's no blocker for this beyond "we should decide to have it or not".

cc @gankro

[aturon]

Nominating for 1.5 FCP discussion.

[alexcrichton]

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

[bluss]

The growing part is a "critical" component where we need to grow & zerofill a vector.

[dhardy]

As bluss said, I'd like either resize or fn ensure_size_is_at_least(&mut self, len: usize, fill: T).

The alternative is cumbersome:

let by = newlen - v.len();
v.extend(iter::repeat(0).take(by));

A constructor to create a zeroed out buffer of predefined length would also be useful (Vec::with_capacity(x) doesn't work with std::io::Read::read()).

[FranklinChen]

I'd like a one-stop way to do what I have to do now:

let mut a = Vec::with_capacity(k);
a.resize(k, x);

[huonw]

vec![x; k] should work for that.

[FranklinChen]

@huonw D'oh, thanks! I honestly spent a long time looking for how to do that and couldn't find it. I just found it at https://doc.rust-lang.org/nightly/std/vec/struct.Vec.html

I'm very concerned that I've been having a lot of trouble finding out this kind of information (another problem was finding slice syntax information), because it's embedded in text rather than displayed as a "method". Perhaps there should be bulleted documentation for all syntax and macros in its own section? Is it my fault I don't necessarily read whole paragraphs of text that are classified as "examples"?

[bluss]

@dhardy The alternative is cumbersome and for now, much less efficient, I'll point out.

[dhardy]

@bluss thanks. I didn't realise I could create a vec with vec![x; k] but would still like resize.

[bluss]

I'm a bit worried by all the people that don't find vec![x; N].

[petrochenkov]

@bluss
Can you reopen #26723 ?
It's still a problem and the motivation for closing it was unconvincing, but I forgot about it at the time.

[steveklabnik]

Yeah, discovery is hard :/

[alexcrichton]

The libs team discussed this today and the decision was to stabilize.

[bluss]

No comment on the dual role of growing and shrinking? Why have both? Only the grow part seems useful (and only the grow part is unique to this method).

[Gankra]

@bluss the dual role was frequently requested. I was personally happy when we just had truncate and whatever the growing version was. shrug

[bluss]

I see.

[Gankra]

fwiw this was accepted in collections reform part 2.

See e.g. rust-lang/rfcs#509 (comment) for discussion