Update documentation for arbitrary_enum_discriminant feature #1055

fee1-dead · 2021-07-04T06:19:48Z

Continuation of #639. This should only be merged after the feature is stabilized.

Stabilize `arbitrary_enum_discriminant` Closes rust-lang#60553. ---- ## Stabilization Report _copied from rust-lang#60553 (comment) ### Summary Enables a user to specify *explicit* discriminants on arbitrary enums. Previously, this was hard to achieve: ```rust #[repr(u8)] enum Foo { A(u8) = 0, B(i8) = 1, C(bool) = 42, } ``` Someone would need to add 41 hidden variants in between as a workaround with implicit discriminants. In conjunction with [RFC 2195](https://github.com/rust-lang/rfcs/blob/master/text/2195-really-tagged-unions.md), this feature would provide more flexibility for FFI and unsafe code involving enums. ### Test cases Most tests are in [`src/test/ui/enum-discriminant`](https://github.com/rust-lang/rust/tree/master/src/test/ui/enum-discriminant), there are two [historical](https://github.com/rust-lang/rust/blob/master/src/test/ui/parser/tag-variant-disr-non-nullary.rs) [tests](https://github.com/rust-lang/rust/blob/master/src/test/ui/parser/issue-17383.rs) that are now covered by the feature (removed by this pr due to them being obsolete). ### Edge cases The feature is well defined and does not have many edge cases. One [edge case](rust-lang#70509) was related to another unstable feature named `repr128` and is resolved. ### Previous PRs The [implementation PR](rust-lang#60732) added documentation to the Unstable Book, rust-lang/reference#1055 was opened as a continuation of rust-lang/reference#639. ### Resolution of unresolved questions The questions are resolved in rust-lang#60553 (comment). ---- (someone please add `needs-fcp`)

ehuss

Thanks, this looks great! 🎉

There's another link in const_eval.md that references #custom-discriminant-values-for-fieldless-enumerations that needs to be updated.

src/items/enumerations.md

ehuss · 2021-08-19T03:24:05Z

src/items/enumerations.md

+<ol>
+<li>
+
+if the enumeration is "C-like" (i.e., it has no tuple or struct variants); e.g.:


Suggested change

if the enumeration is "C-like" (i.e., it has no tuple or struct variants); e.g.:

if the enumeration is fieldless (i.e., it has no tuple or struct variants); e.g.:

This is wrong because the compiler would tell me to add a repr for the enum.

I don't understand what this has to do with the representation. We have been moving the reference to use the term "fieldless enum" instead of "c-like"; that is the terminology we would prefer.

True, but only c-like enums are allowed to have a discriminant without a #[repr()] attribute.

Sorry, I'm still confused. The term "c-like enum" has historically been used for Rust enums that only has unit variants (that is, variants without fields). We have changed the terminology in the reference to use the phrase "fieldless enum" instead. We want to stop using the phrase "c-like" altogether. Indeed, this particular section is showing a fieldless enum. Why do you still want to use the term "c-like" here?

Hm, I can see how the literal interpretation of "fieldless" could be confused with those. The two examples with brackets define tuple and struct variants, they just have 0 fields. The unit variant has a distinctly different meaning which has different semantics. As defined here in the reference, a "fieldless enum" is one with only unit variants (your first example).

I don't think that is correct. The reference says

An enum where no constructors contain fields are called a
field-less enum.

Well, this PR is the one adding that definition, no? I guess my review could be more thorough here, but if you want to make that statement more precise, I would reword it to be that a fieldless enum is one that only has unit variants. I would also drop the hyphen. AFAIK, the language has always generally treated unit variants different from struct and tuple variants, even if the later two don't have fields.

all "fieldless" enums can be cast to their discriminant values. however only c-like enums may specify discriminants without a repr. @jswrenn has already said this on the original PR: #639 (comment)

Opened rust-lang/rust#88203. This should be made consistent and I see no problems for fieldless enums to have an unspecified repr.

src/items/enumerations.md

fee1-dead · 2022-04-06T01:32:33Z

@rustbot label -S-waiting-on-author

rustbot · 2022-04-06T01:32:34Z

Error: The feature shortcut is not enabled in this repository.
To enable it add its section in the triagebot.toml in the root of the repository.

Please let @rust-lang/release know if you're having trouble with this bot.

…ry_enum_discriminant, r=joshtriplett Stabilize arbitrary_enum_discriminant, take 2 Documentation has been updated in rust-lang/reference#1055. cc rust-lang#86860 for previous stabilization report. Not yet marks rust-lang#60553 as done: need documentation in the rust reference.

ehuss

Returning to this now that it has landed. I think the structure and content look great. I just want to nail down the details. Other than the comments I added, is this otherwise up to date with what eventually finalized?

ehuss · 2022-10-28T01:00:20Z

src/items/enumerations.md

+following the variant name with `=` and a [constant expression]:
+
+
+1. if the enumeration is "C-like" (i.e., it has no tuple or struct variants); e.g.:


I feel like we tried quite hard to scrub the presence of the "C-like" terminology, and this PR seems to bring it back. Would it be possible to come up with some other terminology here? Or, perhaps it doesn't need an explicit term, but could just say "if the enumeration only has Unit variants…"?

Yeah, maybe we can use Unit-only and link to a section when referring to it somewhere else. Will do so in a moment.

ehuss · 2022-10-28T01:02:56Z

src/items/enumerations.md

+If an enumeration is C-like (with no tuple and struct variants), then its
+discriminant can be directly accessed with a [numeric cast]; e.g.:


This seems to be missing the grandfathered support for casting enums with empty constructors, like:

enum Fieldless { Tuple(), Struct{}, Unit, } assert_eq!(Fieldless::Tuple() as u8, 0); assert_eq!(Fieldless::Struct{} as u8, 1); assert_eq!(Fieldless::Unit as u8, 2);

Was that excluded for a reason?

The problem is that they cannot be casted to their discriminant if their discriminants are specified. This is summarized in rust-lang/rust#60553 (comment). This will likely need more discussion around the design, but none of the design concerns block the stabilization of arbitrary_enum_discriminant.

If you would like this to be included, then I can put that in.

ehuss · 2022-10-28T01:07:05Z

src/items/enumerations.md

+An enum where no constructors contain fields are called a
+*<a id="field-less-enum">field-less enum</a>*.


I think the field-less term got a little confusing, and it's still not clear to me, does it also cover empty constructors, like:

enum Fieldless { Tuple(), Struct{}, Unit, }

If so, perhaps that needs to be made explicit?

Yeah, I will add that as an example.

Co-Authored-By: Ryan Scheel <Ryan.havvy@gmail.com>

scottmcm · 2022-11-25T21:13:55Z

src/items/enumerations.md

+
+impl Enum {
+    fn discriminant(&self) -> u8 {
+        unsafe { *(self as *const Self as *const u8) }


This doesn't look right to me? Just because it's repr(u8) I don't think that means that the discriminant is necessarily first.

If the enum were repr(C) or repr(C, u8) then it'd be ok because the field order would be specified. But for repr(Rust) I'm pretty sure we're allowed to put the discriminant after the payload, if we wanted to.

Oh, never mind, it looks like https://rust-lang.github.io/rfcs/2195-really-tagged-unions.html#guide-level-explanation means that repr(u8) disables all field reordering too, not just sets the discriminant size :(

I found that odd, too. The RFC justifies this by saying the discriminant position needs to be deterministic, which seems possible to do without disabling other field reordering?

(At this point it seems too late to change, but an edition change could require repr(C) for deterministic field ordering.)

That would need another flag to distinguish between the union-of-structs and struct-containing-union-of-structs layouts, though. Which might be nice, though!

I suppose the argument is that the size of the discriminant can't matter to safe code, so the only point of ever specifying its width is to get deterministic layout?

Explain how to get the discriminant out of a `#[repr(T)] enum` with payload example stolen from rust-lang/reference#1055 `@rustbot` label A-docs

Explain how to get the discriminant out of a `#[repr(T)] enum` with payload example stolen from rust-lang/reference#1055 ``@rustbot`` label A-docs

Explain how to get the discriminant out of a `#[repr(T)] enum` with payload example stolen from rust-lang/reference#1055 ```@rustbot``` label A-docs

Explain how to get the discriminant out of a `#[repr(T)] enum` with payload example stolen from rust-lang/reference#1055 ````@rustbot```` label A-docs

fee1-dead · 2022-11-28T06:56:14Z

@rustbot ready

rustbot · 2022-11-28T06:56:15Z

Error: The feature shortcut is not enabled in this repository.
To enable it add its section in the triagebot.toml in the root of the repository.

Please file an issue on GitHub at triagebot if there's a problem with this bot, or reach out on #t-infra on Zulip.

fee1-dead · 2022-11-28T06:56:35Z

@rustbot label -S-waiting-on-author +S-waiting-on-review

The field-less casting was grandfathered in when arbitrary_enum_discriminant was stabilized. It probably shouldn't be allowed.

The reference only uses uppercase for the first letter.

Although both work, semantically an `<a>` without an `href` is intended to be a placeholder for a link.

This fixes some broken links.

ehuss

I went ahead and pushed some edits to fix some issues, but also to add the grandfathered behavior of allowing as casting for fieldless enums. If anyone feels that it should be different, feel free to open a PR.

ehuss · 2023-01-08T17:31:43Z

Thanks again fee1-dead, and sorry for the super long delay.

Update books ## rust-lang/book 1 commits in 2bd5d42c9956369132228da6409f0e68da56c51a..2cd1b5593d26dc6a03c20f8619187ad4b2485552 2023-01-12 14:47:47 UTC to 2023-01-12 14:47:47 UTC - Typo (rust-lang/book#3457) ## rust-lang/nomicon 2 commits in 8ca261268068d80c0969260fff15199bad87b587..960d610e7f33889a2577f5f17c26f0d5c82b30df 2023-01-06 11:51:41 UTC to 2023-01-05 10:20:31 UTC - vec/raw: Simplify `RawVec::grow` (rust-lang/nomicon#392) - borrow-splitting: Use `take` instead of `replace` (rust-lang/nomicon#391) ## rust-lang/reference 4 commits in 3ae62681ff236d5528ef7c8c28ba7c6b2ecc6731..2cb0ed9ba56360949f492f9866afe8c293f9f9da 2023-01-13 03:16:35 UTC to 2023-01-07 00:08:06 UTC - Update field-expr.md (rust-lang/reference#1318) - Update documentation for arbitrary_enum_discriminant feature (rust-lang/reference#1055) - Add links to definitions of terminology ... (rust-lang/reference#1315) - Enable triagebot shortcuts (rust-lang/reference#1314) ## rust-lang/rust-by-example 2 commits in 8888f9428fe9a48f31de6bd2cef9b9bf80791edc..a9fb7d13eadfcc5f457962731f105b97f9a7474a 2023-01-14 10:25:39 UTC to 2023-01-11 18:25:42 UTC - get_or_insert example: print my_fruit as intended (rust-lang/rust-by-example#1664) - Update print.md (rust-lang/rust-by-example#1663) ## rust-lang/rustc-dev-guide 8 commits in b3e2a6e..7352353 2023-01-14 20:34:23 UTC to 2023-01-02 23:35:09 UTC - fix examples for rustc 1.68.0-nightly (935dc07 2022-12-19) (rust-lang#1556) (rust-lang/rustc-dev-guide#1557) - Update incremental-compilation-in-detail.md (rust-lang/rustc-dev-guide#1553) - Link to the youtube recording of my talk, not the summary (rust-lang/rustc-dev-guide#1554) - Change `src/test` to `tests` (rust-lang/rustc-dev-guide#1547) - add full name for ICE (rust-lang/rustc-dev-guide#1552) - Fix incorrect links (rust-lang/rustc-dev-guide#1549) - fix rebase link (rust-lang/rustc-dev-guide#1546) - Add a section for how to review code more easily (rust-lang/rustc-dev-guide#1538)

Explain how to get the discriminant out of a `#[repr(T)] enum` with payload example stolen from rust-lang/reference#1055 ````@rustbot```` label A-docs

…strieb Update links for custom discriminants. The discriminant documentation was updated in rust-lang/reference#1055 which changed the layout a bit. This updates the links to the updated locations.

Update links for custom discriminants. The discriminant documentation was updated in rust-lang/reference#1055 which changed the layout a bit. This updates the links to the updated locations.

fee1-dead mentioned this pull request Jul 4, 2021

Stabilize arbitrary_enum_discriminant rust-lang/rust#86860

Merged

ehuss added the S-waiting-on-stabilization Waiting for a stabilization PR to be merged in the main Rust repository label Jul 4, 2021

ehuss reviewed Aug 19, 2021

View reviewed changes

ehuss added S-waiting-on-author Status: The marked PR is awaiting some action (such as code changes) from the PR author. and removed S-waiting-on-stabilization Waiting for a stabilization PR to be merged in the main Rust repository labels Aug 19, 2021

fee1-dead force-pushed the arbitrary_enum_discriminant branch from fa1da0b to db8d5b0 Compare August 19, 2021 04:49

fee1-dead requested a review from ehuss August 19, 2021 07:20

fee1-dead mentioned this pull request Aug 21, 2021

Make specifying repr optional for fieldless enums rust-lang/rust#88203

Closed

nikomatsakis mentioned this pull request Nov 30, 2021

Tracking issue for RFC 2363, "Allow arbitrary enums to have explicit discriminants" rust-lang/rust#60553

Closed

5 tasks

fee1-dead force-pushed the arbitrary_enum_discriminant branch from db8d5b0 to 2e93a1f Compare April 6, 2022 01:32

This comment was marked as resolved.

Sign in to view

fee1-dead mentioned this pull request Apr 6, 2022

Stabilize arbitrary_enum_discriminant, take 2 rust-lang/rust#95710

Merged

fee1-dead force-pushed the arbitrary_enum_discriminant branch from 2e93a1f to ccce6d9 Compare April 6, 2022 01:42

ehuss reviewed Oct 28, 2022

View reviewed changes

jswrenn and others added 5 commits October 28, 2022 13:14

Update enum documentation for arbitrary_enum_discriminant feature.

45f85e1

"e.g." -> "for example"

3b5ac6c

Co-Authored-By: Ryan Scheel <Ryan.havvy@gmail.com>

Applied suggestions from Havvy

f8ac109

Apply Suggestions

3d4745b

Update terminology from fieldless to unit-only

5d9300c

fee1-dead force-pushed the arbitrary_enum_discriminant branch from ccce6d9 to 5d9300c Compare October 28, 2022 07:01

ehuss mentioned this pull request Nov 9, 2022

Update documentation for arbitrary_enum_discriminant feature #639

Closed

lukas-code mentioned this pull request Nov 25, 2022

Explain how to get the discriminant out of a #[repr(T)] enum with payload rust-lang/rust#104892

Merged

scottmcm reviewed Nov 25, 2022

View reviewed changes

rustbot added S-waiting-on-review Status: The marked PR is awaiting review from a maintainer and removed S-waiting-on-author Status: The marked PR is awaiting some action (such as code changes) from the PR author. labels Nov 28, 2022

ehuss added 4 commits January 8, 2023 09:06

Clarify exactly which kinds of enums can be as casted.

c351aa7

The field-less casting was grandfathered in when arbitrary_enum_discriminant was stabilized. It probably shouldn't be allowed.

Normalize section header capitalization.

1b38681

The reference only uses uppercase for the first letter.

Use span instead of <a> for anchor placement.

9af0aa4

Although both work, semantically an `<a>` without an `href` is intended to be a placeholder for a link.

Add anchor for removed section.

4f32346

This fixes some broken links.

ehuss approved these changes Jan 8, 2023

View reviewed changes

ehuss merged commit cc07ac0 into rust-lang:master Jan 8, 2023

rustbot mentioned this pull request Jan 16, 2023

Update books rust-lang/rust#106942

Merged

ehuss mentioned this pull request Mar 18, 2023

Update links for custom discriminants. rust-lang/rust#109317

Merged

fee1-dead deleted the arbitrary_enum_discriminant branch March 20, 2024 14:57

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Update documentation for arbitrary_enum_discriminant feature #1055

Update documentation for arbitrary_enum_discriminant feature #1055

fee1-dead commented Jul 4, 2021

ehuss left a comment

ehuss Aug 19, 2021

fee1-dead Aug 19, 2021

ehuss Aug 20, 2021

fee1-dead Aug 20, 2021

ehuss Aug 20, 2021

ehuss Aug 20, 2021

fee1-dead Aug 20, 2021

ehuss Aug 20, 2021

fee1-dead Aug 21, 2021

fee1-dead Aug 21, 2021

fee1-dead commented Apr 6, 2022 •

edited

Loading

rustbot commented Apr 6, 2022

This comment was marked as resolved.

ehuss left a comment

ehuss Oct 28, 2022

fee1-dead Oct 28, 2022

ehuss Oct 28, 2022

fee1-dead Oct 28, 2022

ehuss Oct 28, 2022

fee1-dead Oct 28, 2022

scottmcm Nov 25, 2022

scottmcm Nov 25, 2022

tmandry Dec 2, 2022

tmandry Dec 2, 2022

scottmcm Dec 3, 2022

fee1-dead commented Nov 28, 2022

rustbot commented Nov 28, 2022

fee1-dead commented Nov 28, 2022

ehuss left a comment

ehuss commented Jan 8, 2023

	if the enumeration is "C-like" (i.e., it has no tuple or struct variants); e.g.:
	if the enumeration is fieldless (i.e., it has no tuple or struct variants); e.g.:

		following the variant name with `=` and a [constant expression]:


		1. if the enumeration is "C-like" (i.e., it has no tuple or struct variants); e.g.:

		If an enumeration is C-like (with no tuple and struct variants), then its
		discriminant can be directly accessed with a [numeric cast]; e.g.:

		An enum where no constructors contain fields are called a
		<a id="field-less-enum">field-less enum</a>.

Update documentation for arbitrary_enum_discriminant feature #1055

Update documentation for arbitrary_enum_discriminant feature #1055

Conversation

fee1-dead commented Jul 4, 2021

ehuss left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

fee1-dead commented Apr 6, 2022 • edited Loading

rustbot commented Apr 6, 2022

This comment was marked as resolved.

ehuss left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

fee1-dead commented Nov 28, 2022

rustbot commented Nov 28, 2022

fee1-dead commented Nov 28, 2022

ehuss left a comment

Choose a reason for hiding this comment

ehuss commented Jan 8, 2023

fee1-dead commented Apr 6, 2022 •

edited

Loading