Add equivalents of C's <ctype.h> functions to AsciiExt.

[zackw]

• is_ascii_alphabetic
• is_ascii_uppercase
• is_ascii_lowercase
• is_ascii_alphanumeric
• is_ascii_digit
• is_ascii_hexdigit
• is_ascii_punctuation
• is_ascii_graphic
• is_ascii_whitespace
• is_ascii_control

This addresses issue #39658.

Lightly tested on x86-64-linux. tidy complains about the URLs in the documentation making lines too long, I don't know what to do about that.

[rust-highfive]

Thanks for the pull request, and welcome! The Rust team is excited to review your changes, and you should hear from @sfackler (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 Due to the way GitHub handles out-of-date commits, this should also make it reasonably obvious what issues have or haven't been addressed.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.

[sfackler]

cc @rust-lang/libs

I've always been a bit annoyed this is a separate trait at all. Might be worth inlining these while we're at it.

[zackw]

I've always been a bit annoyed this is a separate trait at all. Might be worth inlining these while we're at it.

I assume by that you mean folding the entire AsciiExt trait into the types it extends. I don't disagree, but can that be done compatibly? There's already a compatibility kludge (the unimplemented!s) in here...

[sfackler]

The idea would be that we deprecate this entire module and just move to inherent methods.

[aturon]

@sfackler

I've always been a bit annoyed this is a separate trait at all. Might be worth inlining these while we're at it.

Interesting -- I don't think there's a good rationale for them being extension methods right now. (Maybe there was at some point, pre-DST). Given that all the methods include ascii in their name, this seems reasonable to me.

I'm also 👍 on this PR.

[zackw]

I don't mind doing the work to turn AsciiExt into inherent methods, but I don't understand how libstd, libcore, and libcollections are organized well enough to know where to put the inherent methods. I'm going to need some help with that.

[aturon]

@japaric, any chance you'd be willing to help @zackw here, given that it may require some more inherent method magic around primitive types?

[alexcrichton]

Looks good to me as well!

I believe the intention of the trait was to add the method to [u8] and u8 as well, as opposed to just char and str.

I'd be tempted to punt the migration to inherent methods (if at all) to a future PR though and go ahead and land this. Thoughts?

[alexcrichton]

@zackw it also looks like tidy may be failing on travis?

[aturon]

I'd be tempted to punt the migration to inherent methods (if at all) to a future PR though and go ahead and land this. Thoughts?

Yeah, that probably makes sense. I do think it'd be a nice change, but it's orthogonal.

[zackw]

@alexcrichton

tidy may be failing on travis

Yeah, that's because the added documentation contains two very long URLs. I have just pushed a change to squeeze those lines in under the limit (by shortening the [foo]: tag for them).

[alexcrichton]

Shame on you for writing thorough documentation!

:)

[zackw]

Okay, now the problem is that the doc tests for the new unstable feature aren't properly annotated to say that they use the new unstable feature. 😝 And the reason I didn't notice this before is that I don't know how to run stdlib doc tests locally. 🤦‍♂️

Can one of you please tell me how to do that?

[alexcrichton]

Oh you can just prefix the tests with #![feature(...)] to allow them to pass. There should be a few other examples of this throughout the standard library as well

[zackw]

Oh you can just prefix the tests with #![feature(...)] to allow them to pass

That part I know. What I don't know is how to run the doc tests locally, so that I can verify that I did it right before pushing. rustdoc --test src/libstd/ascii.rs doesn't work, because it tries to process ascii.rs as a standalone crate.

[alexcrichton]

Oh for that you can use ./x.py src/libstd --test-args $name

[aturon]

We discussed this is the libs team meeting, and we're all in favor of merging as-is. Feel free to open a separate PR to move to inherent methods as well!

r=me once the travis failure is addressed.

[zackw]

@alexcrichton Oh, I see now. x.py test src/libstd does run the doctests, but only If all of the #[test] tests succeed. And my computer can't do IPv6, which makes the #[test] tests always have some failures, so I thought the doctests were under a separate x.py invocation. Thanks for the hint.

[alexcrichton]

@bors: r+

[bors]

📌 Commit 162240c has been approved by alexcrichton

[japaric]

@aturon

@japaric, any chance you'd be willing to help @zackw here, given that it may require some more inherent method magic around primitive types?

Sure!

@zackw For char, u8 and str, I believe the implementations can simply be moved into the inherent impl blocks:

• u8
• char
• str

For [u8], we'll have add a new u8_slice lang time to allow an inherent impl [u8] alongside the generic impl<T> [T] block. The magic for that is around here but I'm not sure if the change will be straightforward. I can look into that part.

[sfackler]

Might be simpler to just stick it in the impl<T> [T] block with a where T: u8 bound on the methods.

[japaric]

with a where T: u8 bound on the methods.

But u8 is a type, not a trait. Or are you suggesting a "double dispatch" approach?

[sfackler]

Oh duh, never mind me.

[SimonSapin]

s/punctuation/graphic/ here.

[SimonSapin]

When doing this, please make sure that [u8] methods show up in rustdoc output.