Implement TryFrom<&[T]> for &[T; N] by nvzqz · Pull Request #44764 · rust-lang/rust

nvzqz · 2017-09-21T20:50:02Z

There are many cases where a buffer with a static compile-time size is preferred over a slice with a dynamic size. This allows for performing a checked conversion from &[T] to &[T; N]. This may also lead to compile-time optimizations involving [T; N] such as loop unrolling.

This is my first PR to Rust, so I'm not sure if discussion of this change should happen here or does it need its own RFC? I figured these changes would be a subset of #33417.

There are many cases where a buffer with a static compile-time size is preferred over a slice with a dynamic size. This allows for performing a checked conversion from &[T] to &[T; N]. This may also lead to compile-time optimizations involving [T; N] such as loop unrolling.

rust-highfive · 2017-09-21T20:50:14Z

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @BurntSushi (or someone else) soon.

If any changes to this PR are deemed necessary, please add them as extra commits. This ensures that the reviewer can see what has changed since they last reviewed the code. Due to the way GitHub handles out-of-date commits, this should also make it reasonably obvious what issues have or haven't been addressed. Large or tricky changes may require several passes of review and changes.

Please see the contribution instructions for more information.

shepmaster · 2017-09-22T16:50:16Z

Thanks for the awesome PR, @nvzqz! We'll wrangle up a top-notch reviewer for you in a short amount of time!

jethrogb · 2017-09-24T17:07:04Z

I'd love to make arrays better, but shouldn't we wait for const generics?

shepmaster · 2017-09-24T17:23:46Z

shouldn't we wait for const generics?

Is there a reason this would conflict or not be implementable with const generics? Otherwise it seems to leave us in no worse shape than we are now.

jethrogb · 2017-09-24T17:29:15Z

Is there a reason this would conflict

Not that I'm aware of. People have previously been cautious about metadata bloat though.

nvzqz · 2017-09-24T20:17:01Z

This change adds 66 new implementations of TryFrom, so I understand the concern about metadata bloat. However, I don't want to wait until integer generics to have this change. It seems likely that TryFrom will be stabilized way before that.

BurntSushi · 2017-09-25T21:24:22Z

I haven't been following TryFrom or const generics, so I'd like to hand this off to someone else. @sfackler?

alexcrichton · 2017-09-28T14:39:42Z

ping @sfackler, mind taking a look at this?

sfackler · 2017-09-28T17:14:26Z

+/// The error type returned when a conversion from a slice to an array fails.
+#[unstable(feature = "try_from", issue = "33417")]
+#[derive(Debug, Copy, Clone)]
+pub struct TryFromSliceError(());


This should probably implement Error in libstd.

Should the Display implementation also be in libstd or libcore?

See e45e8ab

sfackler · 2017-09-28T17:15:16Z

I'm on board with these existing, but let's see if everyone is okay with more ad-hoc array stuff.

@rfcbot fcp merge

sfackler · 2017-09-28T18:13:43Z

Oops, forgot to libs tag it

@rfcbot fcp merge

rfcbot · 2017-09-28T18:52:33Z

Team member @sfackler has proposed to merge this. The next step is review by the rest of the tagged teams:

No concerns currently listed.

Once these reviewers reach consensus, this will enter its final comment period. If you spot a major issue that hasn't been raised at any point in this process, please speak up!

See this document for info about what commands tagged team members can give me.

Enforces use of Error::description instead.

bluss · 2017-09-29T18:23:28Z

slice.len() >= $N would still be valid, but I guess "slicing" to an array shorter than the input slice is not compatible with the spirit of the TryFrom trait (?).

nvzqz · 2017-09-29T18:36:51Z

I believe that the TryFrom conversions should be equal to the length of the slice. You can always do:

let values = [1,2,3,4];let three:&[i32;3] = values[..3].try_into().unwrap();// Or even:let last:&[i32;2] = values[2..4].try_into().unwrap();

sfackler · 2017-09-29T18:38:43Z

In other similar situations like copy_from_slice and clone_from_slice, we require that the lengths match exactly.

bors · 2017-09-30T01:35:57Z

☔ The latest upstream changes (presumably #44174) made this pull request unmergeable. Please resolve the merge conflicts.

carols10cents · 2017-10-09T14:37:19Z

ping @BurntSushi and @aturon for your ticky boxes here!

scottmcm · 2017-10-09T18:29:59Z

If TryFrom starts allowing &[T] -> &]T; N], shouldn't From support &[T; N] -> &[T] for symmetry?

SimonSapin · 2017-10-09T21:42:12Z

&[T; N] already coerce to &[T] implicitly in many places, so unless some specific code could use a From impl in a generic context, that impl doesn’t seem very useful. &[T] to &[T; N] on the other hand is new functionality that previously required unsafe code and macros to achieve.

shepmaster · 2017-10-15T15:45:38Z

ping @aturon for your ticky boxes here!

nvzqz · 2017-10-22T00:09:46Z

Is this just waiting on @aturon? It's been 3 weeks now 😕

carols10cents · 2017-10-23T15:31:08Z

@aturon was out for two weeks, but hopefully should be back now, but he is traveling a bit again this week :) i'll ping him on IRC.

bjorn3 · 2017-10-24T16:01:50Z

+impl fmt::Display for TryFromSliceError {
+ #[inline]
+ fn fmt(&self, f: &mut fmt::Formatter) -> fmt::Result {
+ fmt::Display::fmt(self.__description(), f)


Why not use the description method from the Error trait?

Error is declined in libstd, not libcore

Defined* (I'm on mobile)

I see, never mind.

rfcbot · 2017-10-24T16:26:30Z

🔔 This is now entering its final comment period, as per the review above. 🔔

alexcrichton · 2017-10-31T22:16:53Z

Egad sorry for the delay! This can now be approved after we've all signed off --

@bors: r+

bors · 2017-10-31T22:16:54Z

📌 Commit 4c853ad has been approved by alexcrichton

bors · 2017-10-31T23:06:43Z

⌛ Testing commit 4c853ad with merge f3b900c...

Implement TryFrom<&[T]> for &[T; N] There are many cases where a buffer with a static compile-time size is preferred over a slice with a dynamic size. This allows for performing a checked conversion from `&[T]` to `&[T; N]`. This may also lead to compile-time optimizations involving `[T; N]` such as loop unrolling. This is my first PR to Rust, so I'm not sure if discussion of this change should happen here or does it need its own RFC? I figured these changes would be a subset of #33417.

bors · 2017-11-01T01:45:53Z

☀️ Test successful - status-appveyor, status-travis
Approved by: alexcrichton
Pushing f3b900c to master...

This commit serves to provide the same benefits as PR rust-lang#44764 except for boxes that refer to owned arrays allocated in the heap.

jethrogb · 2019-04-26T19:49:13Z

We're missing impl From<Infallible> for TryFromSliceError, see #60318

rust-highfive assigned BurntSushi Sep 21, 2017

shepmaster added the S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. label Sep 22, 2017

nvzqz changed the title ~~Implement TryFrom for array reference types~~Implement TryFrom<&[T]> for &[T; N]Sep 22, 2017

nvzqz mentioned this pull request Sep 24, 2017
Tracking issue for TryFrom/TryInto traits #33417
Closed
3 tasks

Add TryFrom tests for array reference types
6570e4c

Try to fix array TryFrom tests
a6e70b5

BurntSushi assigned sfackler and unassigned BurntSushi Sep 25, 2017

sfackler reviewed Sep 28, 2017
View reviewed changes

sfackler added the T-libs-api [DEPRECATED; DO NOT USE] label Sep 28, 2017

rfcbot added the proposed-final-comment-period Proposed to merge/close by relevant subteam, see T-<team> label. Will enter FCP once signed off. label Sep 28, 2017

nvzqz mentioned this pull request Sep 29, 2017
Implement TryFrom for CString and CStr #44916
Closed

nvzqz added 4 commits September 29, 2017 11:15

Implement Error trait for TryFromSliceError
e45e8ab

Make TryFromSliceError::__description unstable
e41610c
Enforces use of Error::description instead.

Pass formatter in TryFromSliceError Display impl
d2cf66b

Enable required features for core::array in libstd
d9d8772

Merge remote-tracking branch 'upstream/master'
4c853ad

shepmaster added S-waiting-on-team and removed S-waiting-on-review Status: Awaiting review from the assignee but also interested parties. labels Oct 15, 2017

bjorn3 requested changes Oct 24, 2017
View reviewed changes

rfcbot added the final-comment-period In the final comment period and will be merged soon unless new substantive objections are raised. label Oct 24, 2017

rfcbot removed the proposed-final-comment-period Proposed to merge/close by relevant subteam, see T-<team> label. Will enter FCP once signed off. label Oct 24, 2017

bors merged commit 4c853ad into rust-lang:masterNov 1, 2017

bors mentioned this pull request Nov 1, 2017
Copy all AsciiExt methods to the primitive types directly in order to deprecate it later #44042
Merged

nvzqz added a commit to nvzqz/rust that referenced this pull request Jan 7, 2018
Implement TryFrom for array box types
22c581b
This commit serves to provide the same benefits as PR rust-lang#44764 except for boxes that refer to owned arrays allocated in the heap.

nvzqz mentioned this pull request Jan 7, 2018
Implement TryFrom<Box<[T]>> for Box<[T; N]> #47245
Closed

Uh oh!

Uh oh!

Conversation

nvzqz commented Sep 21, 2017• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

rust-highfive commented Sep 21, 2017

Uh oh!

shepmaster commented Sep 22, 2017

Uh oh!

jethrogb commented Sep 24, 2017

Uh oh!

shepmaster commented Sep 24, 2017

Uh oh!

jethrogb commented Sep 24, 2017

Uh oh!

nvzqz commented Sep 24, 2017

Uh oh!

BurntSushi commented Sep 25, 2017

Uh oh!

alexcrichton commented Sep 28, 2017

Uh oh!

sfacklerSep 28, 2017

Choose a reason for hiding this comment

Uh oh!

nvzqzSep 28, 2017

Choose a reason for hiding this comment

Uh oh!

nvzqzSep 29, 2017

Choose a reason for hiding this comment

Uh oh!

sfackler commented Sep 28, 2017

Uh oh!

sfackler commented Sep 28, 2017

Uh oh!

rfcbot commented Sep 28, 2017• edited by aturon Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

bluss commented Sep 29, 2017

Uh oh!

nvzqz commented Sep 29, 2017• edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

sfackler commented Sep 29, 2017

Uh oh!

bors commented Sep 30, 2017

Uh oh!

carols10cents commented Oct 9, 2017

Uh oh!

scottmcm commented Oct 9, 2017

Uh oh!

SimonSapin commented Oct 9, 2017

Uh oh!

shepmaster commented Oct 15, 2017

Uh oh!

nvzqz commented Oct 22, 2017

Uh oh!

carols10cents commented Oct 23, 2017

Uh oh!

bjorn3Oct 24, 2017

Choose a reason for hiding this comment

Uh oh!

nvzqzOct 24, 2017

Choose a reason for hiding this comment

Uh oh!

nvzqzOct 24, 2017

Choose a reason for hiding this comment

Uh oh!

bjorn3Oct 24, 2017

Choose a reason for hiding this comment

Uh oh!

rfcbot commented Oct 24, 2017

Uh oh!

alexcrichton commented Oct 31, 2017

Uh oh!

bors commented Oct 31, 2017

Uh oh!

bors commented Oct 31, 2017

Uh oh!

bors commented Nov 1, 2017

nvzqz commented Sep 21, 2017•
edited
Loading

rfcbot commented Sep 28, 2017•
edited by aturon
Loading

nvzqz commented Sep 29, 2017•
edited
Loading