Tracking issue for slice_concat_ext stabilization
#27747
Labels
A-slice
Area: `[T]`
B-unstable
Blocker: Implemented in the nightly compiler and unstable.
C-tracking-issue
Category: An issue tracking the progress of sth. like the implementation of an RFC
disposition-close
This PR / issue is in PFCP or FCP with a disposition to close it.
finished-final-comment-period
The final comment period is finished for this PR / Issue.
Libs-Tracked
Libs issues that are tracked on the team's project board.
T-libs-api
Relevant to the library API team, which will review and decide on the PR/issue.
Comments
aturon
added
T-libs-api
|
Having just run into needing this today it would be great to see the documentation for this referenced in the Vec docs with some examples |
|
@BFrog agreed. cc @steveklabnik |
|
/cc #29380 |
|
@aturon just reading along, and I'm curious if you have time to explain: What are the technical reasons you mentioned at the top? Is this related to why the |
|
So, the main problem with moving to inherent methods at this point is that the traits provide something akin to method overloading -- both apply to slice types varying only by some bounds ( re: If you're just using ... Hope that helps! |
|
Good to know :) |
|
It's kinda annoying to be able to join strings with a str (which can have multiple chars), but joining a slice of slices, you can only join with a single element. e.g. let v = vec!["1", "2", "3"];
v.join(", ")works, but let v: Vec<&[u8]> = vec![b"1", b"2", b"3"];
v.join(b", ") // Error: expected u8, found array of 2 elements
v.join(&b',') // works, but is missing the space chardoesn't work |
Mark-Simulacrum
added
the
C-tracking-issue
|
I tried making the methods inherent with separate inherent impls. But we can't have multiple inherent impls for a primitive. I got the error: "only a single inherent implementation marked with This is known of course, the issue is #32631. |
|
@leodasvacas I think you mean multiple The libs team discussed this and consensus was to add Design proposals to generalize this functionality to iterators are welcome. |
Does this mean |
|
@sanmai-NL That is correct. (Modulo #15702, which is a bug and not considered part of the stability promise.) |
|
Does there exist a path to stabilizing impl<S: Borrow<BStr>> SliceConcatExt for [S] {
type Output = BString;
// ...
}The work-around is to define a separate trait that is probably identical to |
|
Ping @aturon |
|
Is there a reason this trait isn't implemented for |
|
Is there any stable way to use join in a no_std enviroment using alloc? |
|
@nico-abram I believe there is none today, since liballoc does not have a prelude that gets implicitly imported. I a previous comment, I wrote:
This is however an improvement in that the new helper trait does not need to be in scope for the inherent methods to be used. In particular, it does not need to be in the prelude. #62403 implements this. |
@BurntSushi, impl coherence rules do not allow this impl outside of the crate that defines the trait: You would need separate impls
FWIW, not having this benefit is rather typical of functionality defined outside of the standard library. |
Centril
added a commit
to Centril/rust
that referenced
this issue
Jul 22, 2019
Add joining slices of slices with a slice separator, not just a single item rust-lang#27747 (comment) > It's kinda annoying to be able to join strings with a str (which can have multiple chars), but joining a slice of slices, you can only join with a single element. This turns out to be fixable, with some possible inference regressions. # TL;DR Related trait(s) are unstable and tracked at rust-lang#27747, but the `[T]::join` method that is being extended here is already stable. Example use of the new insta-stable functionality: ```rust let nested: Vec<Vec<Foo>> = /* … */; let separator: &[Foo] = /* … */; // Previously: could only be a single &Foo nested.join(separator) ``` Complete API affected by this PR, after changes: ```rust impl<T> [T] { pub fn concat<Item: ?Sized>(&self) -> <Self as Concat<Item>>::Output where Self: Concat<Item> { Concat::concat(self) } pub fn join<Separator>(&self, sep: Separator) -> <Self as Join<Separator>>::Output where Self: Join<Separator> { Join::join(self, sep) } } // The `Item` parameter is only useful for the the slice-of-slices impl. pub trait Concat<Item: ?Sized> { type Output; fn concat(slice: &Self) -> Self::Output; } pub trait Join<Separator> { type Output; fn join(slice: &Self, sep: Separator) -> Self::Output; } impl<T: Clone, V: Borrow<[T]>> Concat<T> for [V] { type Output = Vec<T>; } impl<T: Clone, V: Borrow<[T]>> Join<&'_ T> for [V] { type Output = Vec<T>; } // New functionality here! impl<T: Clone, V: Borrow<[T]>> Join<&'_ [T]> for [V] { type Output = Vec<T>; } impl<S: Borrow<str>> Concat<str> for [S] { type Output = String; } impl<S: Borrow<str>> Join<&'_ str> for [S] { type Output = String; } ``` # Details After rust-lang#62403 but before this PR, the API is: ```rust impl<T> [T] { pub fn concat<Separator: ?Sized>(&self) -> T::Output where T: SliceConcat<Separator> { SliceConcat::concat(self) } pub fn join<Separator: ?Sized>(&self, sep: &Separator) -> T::Output where T: SliceConcat<Separator> { SliceConcat::join(self, sep) } } pub trait SliceConcat<Separator: ?Sized>: Sized { type Output; fn concat(slice: &[Self]) -> Self::Output; fn join(slice: &[Self], sep: &Separator) -> Self::Output; } impl<T: Clone, V: Borrow<[T]>> SliceConcat<T> for V { type Output = Vec<T>; } impl<S: Borrow<str>> SliceConcat<str> for S { type Output = String; } ``` By adding a trait impl we should be able to accept a slice of `T` as the separator, as an alternative to a single `T` value. In a `some_slice.join(some_separator)` call, trait resolution will pick an impl or the other based on the type of `some_separator`. In `some_slice.concat()` however there is no separator, so this call would become ambiguous. Some regression in type inference or trait resolution may be acceptable on principle, but requiring a turbofish for every single call to `concat` isn’t great. The solution to that is splitting the `SliceConcat` trait into two `Concat` and `Join` traits, one for each eponymous method. Only `Join` would gain a new impl, so that `some_slice.concat()` would not become ambiguous. Now, at the trait level the `Concat` trait does not need a `Separator` parameter anymore. However, simply removing it causes one of the impls not to be accepted anymore: ```rust error[E0207]: the type parameter `T` is not constrained by the impl trait, self type, or predicates --> src/liballoc/slice.rs:608:6 | 608 | impl<T: Clone, V: Borrow<[T]>> Concat for [V] { | ^ unconstrained type parameter ``` This makes sense: if `[V]::concat` is a method that is itself not generic, then its return type (which is the `Concat::Output` associated type) needs to be determined based on solely `V`. And although there is no such type in the standard library, there is nothing stopping another crate from defining a `V` type that implements both `Borrow<[Foo]>` and `Borrow<[Bar]>`. It might not be a good idea, but it’s possible. Both would apply here, and there would be no way to determine `T`. This could be a warning sign that this API is too generic. Perhaps we’d be better off having one less type variable, and only implement `Concat for [&'_ [T]]` and `Concat for [Vec<T>]` etc. However this aspect of `[V]::concat` is already stable, so we’re stuck with it. The solution is to keep a dummy type parameter on the `Concat` trait. That way, if a type has multiple `Borrow<[_]>` impls, it’ll end up with multiple corresponding `Concat<_>` impls. In `impl<S: Borrow<str>> Concat<str> for [S]`, the second occurrence of `str` is not meaningful. It could be any type. As long as there is only once such type with an applicable impl, trait resolution will be appeased without demanding turbofishes. # Joining strings with `char` For symmetry I also tried adding this impl (because why not): ```rust impl<S: Borrow<str>> Join<char> for [S] { type Output = String; } ``` This immediately caused an inference regression in a dependency of rustc: ```rust error[E0277]: the trait bound `std::string::String: std::borrow::Borrow<[std::string::String]>` is not satisfied --> /home/simon/.cargo/registry/src/github.com-1ecc6299db9ec823/getopts-0.2.19/src/lib.rs:595:37 | 595 | row.push_str(&desc_rows.join(&desc_sep)); | ^^^^ the trait `std::borrow::Borrow<[std::string::String]>` is not implemented for `std::string::String` | = help: the following implementations were found: <std::string::String as std::borrow::Borrow<str>> = note: required because of the requirements on the impl of `std::slice::Join<&std::string::String>` for `[std::string::String]` ``` In the context of this code, two facts are known: * `desc_rows` is a `Vec<String>` * `desc_sep` is a `String` Previously the first fact alone reduces the resolution of `join` to only one solution, where its argument it expected to be `&str`. Then, `&String` is coerced to `&str`. With the new `Join` impl, the first fact leavs two applicable impls where the separator can be either `&str` or `char`. But `&String` is neither of these things. It appears that possible coercions are not accounted for, in the search for a solution in trait resolution. I have not included this new impl in this PR. It’s still possible to add later, but the `getopts` breakage does not need to block the rest of the PR. And the functionality easy for end-user to duplicate: `slice_of_strings.join(&*char_separator.encode_utf8(&mut [0_u8, 4]))` The `&*` part of that last code snippet is another case of the same issue: `encode_utf8` returns `&mut str` which can be coerced to `&str`, but isn’t when trait resolution is ambiguous.
Centril
added a commit
to Centril/rust
that referenced
this issue
Jul 23, 2019
Add joining slices of slices with a slice separator, not just a single item rust-lang#27747 (comment) > It's kinda annoying to be able to join strings with a str (which can have multiple chars), but joining a slice of slices, you can only join with a single element. This turns out to be fixable, with some possible inference regressions. # TL;DR Related trait(s) are unstable and tracked at rust-lang#27747, but the `[T]::join` method that is being extended here is already stable. Example use of the new insta-stable functionality: ```rust let nested: Vec<Vec<Foo>> = /* … */; let separator: &[Foo] = /* … */; // Previously: could only be a single &Foo nested.join(separator) ``` Complete API affected by this PR, after changes: ```rust impl<T> [T] { pub fn concat<Item: ?Sized>(&self) -> <Self as Concat<Item>>::Output where Self: Concat<Item> { Concat::concat(self) } pub fn join<Separator>(&self, sep: Separator) -> <Self as Join<Separator>>::Output where Self: Join<Separator> { Join::join(self, sep) } } // The `Item` parameter is only useful for the the slice-of-slices impl. pub trait Concat<Item: ?Sized> { type Output; fn concat(slice: &Self) -> Self::Output; } pub trait Join<Separator> { type Output; fn join(slice: &Self, sep: Separator) -> Self::Output; } impl<T: Clone, V: Borrow<[T]>> Concat<T> for [V] { type Output = Vec<T>; } impl<T: Clone, V: Borrow<[T]>> Join<&'_ T> for [V] { type Output = Vec<T>; } // New functionality here! impl<T: Clone, V: Borrow<[T]>> Join<&'_ [T]> for [V] { type Output = Vec<T>; } impl<S: Borrow<str>> Concat<str> for [S] { type Output = String; } impl<S: Borrow<str>> Join<&'_ str> for [S] { type Output = String; } ``` # Details After rust-lang#62403 but before this PR, the API is: ```rust impl<T> [T] { pub fn concat<Separator: ?Sized>(&self) -> T::Output where T: SliceConcat<Separator> { SliceConcat::concat(self) } pub fn join<Separator: ?Sized>(&self, sep: &Separator) -> T::Output where T: SliceConcat<Separator> { SliceConcat::join(self, sep) } } pub trait SliceConcat<Separator: ?Sized>: Sized { type Output; fn concat(slice: &[Self]) -> Self::Output; fn join(slice: &[Self], sep: &Separator) -> Self::Output; } impl<T: Clone, V: Borrow<[T]>> SliceConcat<T> for V { type Output = Vec<T>; } impl<S: Borrow<str>> SliceConcat<str> for S { type Output = String; } ``` By adding a trait impl we should be able to accept a slice of `T` as the separator, as an alternative to a single `T` value. In a `some_slice.join(some_separator)` call, trait resolution will pick an impl or the other based on the type of `some_separator`. In `some_slice.concat()` however there is no separator, so this call would become ambiguous. Some regression in type inference or trait resolution may be acceptable on principle, but requiring a turbofish for every single call to `concat` isn’t great. The solution to that is splitting the `SliceConcat` trait into two `Concat` and `Join` traits, one for each eponymous method. Only `Join` would gain a new impl, so that `some_slice.concat()` would not become ambiguous. Now, at the trait level the `Concat` trait does not need a `Separator` parameter anymore. However, simply removing it causes one of the impls not to be accepted anymore: ```rust error[E0207]: the type parameter `T` is not constrained by the impl trait, self type, or predicates --> src/liballoc/slice.rs:608:6 | 608 | impl<T: Clone, V: Borrow<[T]>> Concat for [V] { | ^ unconstrained type parameter ``` This makes sense: if `[V]::concat` is a method that is itself not generic, then its return type (which is the `Concat::Output` associated type) needs to be determined based on solely `V`. And although there is no such type in the standard library, there is nothing stopping another crate from defining a `V` type that implements both `Borrow<[Foo]>` and `Borrow<[Bar]>`. It might not be a good idea, but it’s possible. Both would apply here, and there would be no way to determine `T`. This could be a warning sign that this API is too generic. Perhaps we’d be better off having one less type variable, and only implement `Concat for [&'_ [T]]` and `Concat for [Vec<T>]` etc. However this aspect of `[V]::concat` is already stable, so we’re stuck with it. The solution is to keep a dummy type parameter on the `Concat` trait. That way, if a type has multiple `Borrow<[_]>` impls, it’ll end up with multiple corresponding `Concat<_>` impls. In `impl<S: Borrow<str>> Concat<str> for [S]`, the second occurrence of `str` is not meaningful. It could be any type. As long as there is only once such type with an applicable impl, trait resolution will be appeased without demanding turbofishes. # Joining strings with `char` For symmetry I also tried adding this impl (because why not): ```rust impl<S: Borrow<str>> Join<char> for [S] { type Output = String; } ``` This immediately caused an inference regression in a dependency of rustc: ```rust error[E0277]: the trait bound `std::string::String: std::borrow::Borrow<[std::string::String]>` is not satisfied --> /home/simon/.cargo/registry/src/github.com-1ecc6299db9ec823/getopts-0.2.19/src/lib.rs:595:37 | 595 | row.push_str(&desc_rows.join(&desc_sep)); | ^^^^ the trait `std::borrow::Borrow<[std::string::String]>` is not implemented for `std::string::String` | = help: the following implementations were found: <std::string::String as std::borrow::Borrow<str>> = note: required because of the requirements on the impl of `std::slice::Join<&std::string::String>` for `[std::string::String]` ``` In the context of this code, two facts are known: * `desc_rows` is a `Vec<String>` * `desc_sep` is a `String` Previously the first fact alone reduces the resolution of `join` to only one solution, where its argument it expected to be `&str`. Then, `&String` is coerced to `&str`. With the new `Join` impl, the first fact leavs two applicable impls where the separator can be either `&str` or `char`. But `&String` is neither of these things. It appears that possible coercions are not accounted for, in the search for a solution in trait resolution. I have not included this new impl in this PR. It’s still possible to add later, but the `getopts` breakage does not need to block the rest of the PR. And the functionality easy for end-user to duplicate: `slice_of_strings.join(&*char_separator.encode_utf8(&mut [0_u8, 4]))` The `&*` part of that last code snippet is another case of the same issue: `encode_utf8` returns `&mut str` which can be coerced to `&str`, but isn’t when trait resolution is ambiguous.
Centril
added a commit
to Centril/rust
that referenced
this issue
Jul 24, 2019
Add joining slices of slices with a slice separator, not just a single item rust-lang#27747 (comment) > It's kinda annoying to be able to join strings with a str (which can have multiple chars), but joining a slice of slices, you can only join with a single element. This turns out to be fixable, with some possible inference regressions. # TL;DR Related trait(s) are unstable and tracked at rust-lang#27747, but the `[T]::join` method that is being extended here is already stable. Example use of the new insta-stable functionality: ```rust let nested: Vec<Vec<Foo>> = /* … */; let separator: &[Foo] = /* … */; // Previously: could only be a single &Foo nested.join(separator) ``` Complete API affected by this PR, after changes: ```rust impl<T> [T] { pub fn concat<Item: ?Sized>(&self) -> <Self as Concat<Item>>::Output where Self: Concat<Item> { Concat::concat(self) } pub fn join<Separator>(&self, sep: Separator) -> <Self as Join<Separator>>::Output where Self: Join<Separator> { Join::join(self, sep) } } // The `Item` parameter is only useful for the the slice-of-slices impl. pub trait Concat<Item: ?Sized> { type Output; fn concat(slice: &Self) -> Self::Output; } pub trait Join<Separator> { type Output; fn join(slice: &Self, sep: Separator) -> Self::Output; } impl<T: Clone, V: Borrow<[T]>> Concat<T> for [V] { type Output = Vec<T>; } impl<T: Clone, V: Borrow<[T]>> Join<&'_ T> for [V] { type Output = Vec<T>; } // New functionality here! impl<T: Clone, V: Borrow<[T]>> Join<&'_ [T]> for [V] { type Output = Vec<T>; } impl<S: Borrow<str>> Concat<str> for [S] { type Output = String; } impl<S: Borrow<str>> Join<&'_ str> for [S] { type Output = String; } ``` # Details After rust-lang#62403 but before this PR, the API is: ```rust impl<T> [T] { pub fn concat<Separator: ?Sized>(&self) -> T::Output where T: SliceConcat<Separator> { SliceConcat::concat(self) } pub fn join<Separator: ?Sized>(&self, sep: &Separator) -> T::Output where T: SliceConcat<Separator> { SliceConcat::join(self, sep) } } pub trait SliceConcat<Separator: ?Sized>: Sized { type Output; fn concat(slice: &[Self]) -> Self::Output; fn join(slice: &[Self], sep: &Separator) -> Self::Output; } impl<T: Clone, V: Borrow<[T]>> SliceConcat<T> for V { type Output = Vec<T>; } impl<S: Borrow<str>> SliceConcat<str> for S { type Output = String; } ``` By adding a trait impl we should be able to accept a slice of `T` as the separator, as an alternative to a single `T` value. In a `some_slice.join(some_separator)` call, trait resolution will pick an impl or the other based on the type of `some_separator`. In `some_slice.concat()` however there is no separator, so this call would become ambiguous. Some regression in type inference or trait resolution may be acceptable on principle, but requiring a turbofish for every single call to `concat` isn’t great. The solution to that is splitting the `SliceConcat` trait into two `Concat` and `Join` traits, one for each eponymous method. Only `Join` would gain a new impl, so that `some_slice.concat()` would not become ambiguous. Now, at the trait level the `Concat` trait does not need a `Separator` parameter anymore. However, simply removing it causes one of the impls not to be accepted anymore: ```rust error[E0207]: the type parameter `T` is not constrained by the impl trait, self type, or predicates --> src/liballoc/slice.rs:608:6 | 608 | impl<T: Clone, V: Borrow<[T]>> Concat for [V] { | ^ unconstrained type parameter ``` This makes sense: if `[V]::concat` is a method that is itself not generic, then its return type (which is the `Concat::Output` associated type) needs to be determined based on solely `V`. And although there is no such type in the standard library, there is nothing stopping another crate from defining a `V` type that implements both `Borrow<[Foo]>` and `Borrow<[Bar]>`. It might not be a good idea, but it’s possible. Both would apply here, and there would be no way to determine `T`. This could be a warning sign that this API is too generic. Perhaps we’d be better off having one less type variable, and only implement `Concat for [&'_ [T]]` and `Concat for [Vec<T>]` etc. However this aspect of `[V]::concat` is already stable, so we’re stuck with it. The solution is to keep a dummy type parameter on the `Concat` trait. That way, if a type has multiple `Borrow<[_]>` impls, it’ll end up with multiple corresponding `Concat<_>` impls. In `impl<S: Borrow<str>> Concat<str> for [S]`, the second occurrence of `str` is not meaningful. It could be any type. As long as there is only once such type with an applicable impl, trait resolution will be appeased without demanding turbofishes. # Joining strings with `char` For symmetry I also tried adding this impl (because why not): ```rust impl<S: Borrow<str>> Join<char> for [S] { type Output = String; } ``` This immediately caused an inference regression in a dependency of rustc: ```rust error[E0277]: the trait bound `std::string::String: std::borrow::Borrow<[std::string::String]>` is not satisfied --> /home/simon/.cargo/registry/src/github.com-1ecc6299db9ec823/getopts-0.2.19/src/lib.rs:595:37 | 595 | row.push_str(&desc_rows.join(&desc_sep)); | ^^^^ the trait `std::borrow::Borrow<[std::string::String]>` is not implemented for `std::string::String` | = help: the following implementations were found: <std::string::String as std::borrow::Borrow<str>> = note: required because of the requirements on the impl of `std::slice::Join<&std::string::String>` for `[std::string::String]` ``` In the context of this code, two facts are known: * `desc_rows` is a `Vec<String>` * `desc_sep` is a `String` Previously the first fact alone reduces the resolution of `join` to only one solution, where its argument it expected to be `&str`. Then, `&String` is coerced to `&str`. With the new `Join` impl, the first fact leavs two applicable impls where the separator can be either `&str` or `char`. But `&String` is neither of these things. It appears that possible coercions are not accounted for, in the search for a solution in trait resolution. I have not included this new impl in this PR. It’s still possible to add later, but the `getopts` breakage does not need to block the rest of the PR. And the functionality easy for end-user to duplicate: `slice_of_strings.join(&*char_separator.encode_utf8(&mut [0_u8, 4]))` The `&*` part of that last code snippet is another case of the same issue: `encode_utf8` returns `&mut str` which can be coerced to `&str`, but isn’t when trait resolution is ambiguous.
Centril
added a commit
to Centril/rust
that referenced
this issue
Jul 24, 2019
Add joining slices of slices with a slice separator, not just a single item rust-lang#27747 (comment) > It's kinda annoying to be able to join strings with a str (which can have multiple chars), but joining a slice of slices, you can only join with a single element. This turns out to be fixable, with some possible inference regressions. # TL;DR Related trait(s) are unstable and tracked at rust-lang#27747, but the `[T]::join` method that is being extended here is already stable. Example use of the new insta-stable functionality: ```rust let nested: Vec<Vec<Foo>> = /* … */; let separator: &[Foo] = /* … */; // Previously: could only be a single &Foo nested.join(separator) ``` Complete API affected by this PR, after changes: ```rust impl<T> [T] { pub fn concat<Item: ?Sized>(&self) -> <Self as Concat<Item>>::Output where Self: Concat<Item> { Concat::concat(self) } pub fn join<Separator>(&self, sep: Separator) -> <Self as Join<Separator>>::Output where Self: Join<Separator> { Join::join(self, sep) } } // The `Item` parameter is only useful for the the slice-of-slices impl. pub trait Concat<Item: ?Sized> { type Output; fn concat(slice: &Self) -> Self::Output; } pub trait Join<Separator> { type Output; fn join(slice: &Self, sep: Separator) -> Self::Output; } impl<T: Clone, V: Borrow<[T]>> Concat<T> for [V] { type Output = Vec<T>; } impl<T: Clone, V: Borrow<[T]>> Join<&'_ T> for [V] { type Output = Vec<T>; } // New functionality here! impl<T: Clone, V: Borrow<[T]>> Join<&'_ [T]> for [V] { type Output = Vec<T>; } impl<S: Borrow<str>> Concat<str> for [S] { type Output = String; } impl<S: Borrow<str>> Join<&'_ str> for [S] { type Output = String; } ``` # Details After rust-lang#62403 but before this PR, the API is: ```rust impl<T> [T] { pub fn concat<Separator: ?Sized>(&self) -> T::Output where T: SliceConcat<Separator> { SliceConcat::concat(self) } pub fn join<Separator: ?Sized>(&self, sep: &Separator) -> T::Output where T: SliceConcat<Separator> { SliceConcat::join(self, sep) } } pub trait SliceConcat<Separator: ?Sized>: Sized { type Output; fn concat(slice: &[Self]) -> Self::Output; fn join(slice: &[Self], sep: &Separator) -> Self::Output; } impl<T: Clone, V: Borrow<[T]>> SliceConcat<T> for V { type Output = Vec<T>; } impl<S: Borrow<str>> SliceConcat<str> for S { type Output = String; } ``` By adding a trait impl we should be able to accept a slice of `T` as the separator, as an alternative to a single `T` value. In a `some_slice.join(some_separator)` call, trait resolution will pick an impl or the other based on the type of `some_separator`. In `some_slice.concat()` however there is no separator, so this call would become ambiguous. Some regression in type inference or trait resolution may be acceptable on principle, but requiring a turbofish for every single call to `concat` isn’t great. The solution to that is splitting the `SliceConcat` trait into two `Concat` and `Join` traits, one for each eponymous method. Only `Join` would gain a new impl, so that `some_slice.concat()` would not become ambiguous. Now, at the trait level the `Concat` trait does not need a `Separator` parameter anymore. However, simply removing it causes one of the impls not to be accepted anymore: ```rust error[E0207]: the type parameter `T` is not constrained by the impl trait, self type, or predicates --> src/liballoc/slice.rs:608:6 | 608 | impl<T: Clone, V: Borrow<[T]>> Concat for [V] { | ^ unconstrained type parameter ``` This makes sense: if `[V]::concat` is a method that is itself not generic, then its return type (which is the `Concat::Output` associated type) needs to be determined based on solely `V`. And although there is no such type in the standard library, there is nothing stopping another crate from defining a `V` type that implements both `Borrow<[Foo]>` and `Borrow<[Bar]>`. It might not be a good idea, but it’s possible. Both would apply here, and there would be no way to determine `T`. This could be a warning sign that this API is too generic. Perhaps we’d be better off having one less type variable, and only implement `Concat for [&'_ [T]]` and `Concat for [Vec<T>]` etc. However this aspect of `[V]::concat` is already stable, so we’re stuck with it. The solution is to keep a dummy type parameter on the `Concat` trait. That way, if a type has multiple `Borrow<[_]>` impls, it’ll end up with multiple corresponding `Concat<_>` impls. In `impl<S: Borrow<str>> Concat<str> for [S]`, the second occurrence of `str` is not meaningful. It could be any type. As long as there is only once such type with an applicable impl, trait resolution will be appeased without demanding turbofishes. # Joining strings with `char` For symmetry I also tried adding this impl (because why not): ```rust impl<S: Borrow<str>> Join<char> for [S] { type Output = String; } ``` This immediately caused an inference regression in a dependency of rustc: ```rust error[E0277]: the trait bound `std::string::String: std::borrow::Borrow<[std::string::String]>` is not satisfied --> /home/simon/.cargo/registry/src/github.com-1ecc6299db9ec823/getopts-0.2.19/src/lib.rs:595:37 | 595 | row.push_str(&desc_rows.join(&desc_sep)); | ^^^^ the trait `std::borrow::Borrow<[std::string::String]>` is not implemented for `std::string::String` | = help: the following implementations were found: <std::string::String as std::borrow::Borrow<str>> = note: required because of the requirements on the impl of `std::slice::Join<&std::string::String>` for `[std::string::String]` ``` In the context of this code, two facts are known: * `desc_rows` is a `Vec<String>` * `desc_sep` is a `String` Previously the first fact alone reduces the resolution of `join` to only one solution, where its argument it expected to be `&str`. Then, `&String` is coerced to `&str`. With the new `Join` impl, the first fact leavs two applicable impls where the separator can be either `&str` or `char`. But `&String` is neither of these things. It appears that possible coercions are not accounted for, in the search for a solution in trait resolution. I have not included this new impl in this PR. It’s still possible to add later, but the `getopts` breakage does not need to block the rest of the PR. And the functionality easy for end-user to duplicate: `slice_of_strings.join(&*char_separator.encode_utf8(&mut [0_u8, 4]))` The `&*` part of that last code snippet is another case of the same issue: `encode_utf8` returns `&mut str` which can be coerced to `&str`, but isn’t when trait resolution is ambiguous.
Centril
added a commit
to Centril/rust
that referenced
this issue
Jul 24, 2019
Add joining slices of slices with a slice separator, not just a single item rust-lang#27747 (comment) > It's kinda annoying to be able to join strings with a str (which can have multiple chars), but joining a slice of slices, you can only join with a single element. This turns out to be fixable, with some possible inference regressions. # TL;DR Related trait(s) are unstable and tracked at rust-lang#27747, but the `[T]::join` method that is being extended here is already stable. Example use of the new insta-stable functionality: ```rust let nested: Vec<Vec<Foo>> = /* … */; let separator: &[Foo] = /* … */; // Previously: could only be a single &Foo nested.join(separator) ``` Complete API affected by this PR, after changes: ```rust impl<T> [T] { pub fn concat<Item: ?Sized>(&self) -> <Self as Concat<Item>>::Output where Self: Concat<Item> { Concat::concat(self) } pub fn join<Separator>(&self, sep: Separator) -> <Self as Join<Separator>>::Output where Self: Join<Separator> { Join::join(self, sep) } } // The `Item` parameter is only useful for the the slice-of-slices impl. pub trait Concat<Item: ?Sized> { type Output; fn concat(slice: &Self) -> Self::Output; } pub trait Join<Separator> { type Output; fn join(slice: &Self, sep: Separator) -> Self::Output; } impl<T: Clone, V: Borrow<[T]>> Concat<T> for [V] { type Output = Vec<T>; } impl<T: Clone, V: Borrow<[T]>> Join<&'_ T> for [V] { type Output = Vec<T>; } // New functionality here! impl<T: Clone, V: Borrow<[T]>> Join<&'_ [T]> for [V] { type Output = Vec<T>; } impl<S: Borrow<str>> Concat<str> for [S] { type Output = String; } impl<S: Borrow<str>> Join<&'_ str> for [S] { type Output = String; } ``` # Details After rust-lang#62403 but before this PR, the API is: ```rust impl<T> [T] { pub fn concat<Separator: ?Sized>(&self) -> T::Output where T: SliceConcat<Separator> { SliceConcat::concat(self) } pub fn join<Separator: ?Sized>(&self, sep: &Separator) -> T::Output where T: SliceConcat<Separator> { SliceConcat::join(self, sep) } } pub trait SliceConcat<Separator: ?Sized>: Sized { type Output; fn concat(slice: &[Self]) -> Self::Output; fn join(slice: &[Self], sep: &Separator) -> Self::Output; } impl<T: Clone, V: Borrow<[T]>> SliceConcat<T> for V { type Output = Vec<T>; } impl<S: Borrow<str>> SliceConcat<str> for S { type Output = String; } ``` By adding a trait impl we should be able to accept a slice of `T` as the separator, as an alternative to a single `T` value. In a `some_slice.join(some_separator)` call, trait resolution will pick an impl or the other based on the type of `some_separator`. In `some_slice.concat()` however there is no separator, so this call would become ambiguous. Some regression in type inference or trait resolution may be acceptable on principle, but requiring a turbofish for every single call to `concat` isn’t great. The solution to that is splitting the `SliceConcat` trait into two `Concat` and `Join` traits, one for each eponymous method. Only `Join` would gain a new impl, so that `some_slice.concat()` would not become ambiguous. Now, at the trait level the `Concat` trait does not need a `Separator` parameter anymore. However, simply removing it causes one of the impls not to be accepted anymore: ```rust error[E0207]: the type parameter `T` is not constrained by the impl trait, self type, or predicates --> src/liballoc/slice.rs:608:6 | 608 | impl<T: Clone, V: Borrow<[T]>> Concat for [V] { | ^ unconstrained type parameter ``` This makes sense: if `[V]::concat` is a method that is itself not generic, then its return type (which is the `Concat::Output` associated type) needs to be determined based on solely `V`. And although there is no such type in the standard library, there is nothing stopping another crate from defining a `V` type that implements both `Borrow<[Foo]>` and `Borrow<[Bar]>`. It might not be a good idea, but it’s possible. Both would apply here, and there would be no way to determine `T`. This could be a warning sign that this API is too generic. Perhaps we’d be better off having one less type variable, and only implement `Concat for [&'_ [T]]` and `Concat for [Vec<T>]` etc. However this aspect of `[V]::concat` is already stable, so we’re stuck with it. The solution is to keep a dummy type parameter on the `Concat` trait. That way, if a type has multiple `Borrow<[_]>` impls, it’ll end up with multiple corresponding `Concat<_>` impls. In `impl<S: Borrow<str>> Concat<str> for [S]`, the second occurrence of `str` is not meaningful. It could be any type. As long as there is only once such type with an applicable impl, trait resolution will be appeased without demanding turbofishes. # Joining strings with `char` For symmetry I also tried adding this impl (because why not): ```rust impl<S: Borrow<str>> Join<char> for [S] { type Output = String; } ``` This immediately caused an inference regression in a dependency of rustc: ```rust error[E0277]: the trait bound `std::string::String: std::borrow::Borrow<[std::string::String]>` is not satisfied --> /home/simon/.cargo/registry/src/github.com-1ecc6299db9ec823/getopts-0.2.19/src/lib.rs:595:37 | 595 | row.push_str(&desc_rows.join(&desc_sep)); | ^^^^ the trait `std::borrow::Borrow<[std::string::String]>` is not implemented for `std::string::String` | = help: the following implementations were found: <std::string::String as std::borrow::Borrow<str>> = note: required because of the requirements on the impl of `std::slice::Join<&std::string::String>` for `[std::string::String]` ``` In the context of this code, two facts are known: * `desc_rows` is a `Vec<String>` * `desc_sep` is a `String` Previously the first fact alone reduces the resolution of `join` to only one solution, where its argument it expected to be `&str`. Then, `&String` is coerced to `&str`. With the new `Join` impl, the first fact leavs two applicable impls where the separator can be either `&str` or `char`. But `&String` is neither of these things. It appears that possible coercions are not accounted for, in the search for a solution in trait resolution. I have not included this new impl in this PR. It’s still possible to add later, but the `getopts` breakage does not need to block the rest of the PR. And the functionality easy for end-user to duplicate: `slice_of_strings.join(&*char_separator.encode_utf8(&mut [0_u8, 4]))` The `&*` part of that last code snippet is another case of the same issue: `encode_utf8` returns `&mut str` which can be coerced to `&str`, but isn’t when trait resolution is ambiguous.
|
Well I’m getting that So it maybe not possible? I guess I should I take this somewhere else if it is. |
|
Interesting, I would have thought those wouldn’t conflict since So yeah, it looks like this impl is not possible without changing trait coherence rules. |
|
Just a thought. Would it be possible to work in something like: format!("Here we are: {}", big_array.lazy_join(", "));where |
|
Yeah that sounds possible. It’s a different feature though. And one you could implement on crates.io with an extension trait. |
|
@camsteffen you might be looking for |
KodrAus
added
A-slice
|
We discussed this in the libs team and the consensus is that we will likely never stabilize the @rfcbot close |
|
Team member @Amanieu has proposed to close this. The next step is review by the rest of the tagged team members: No concerns currently listed. Once a majority of reviewers approve (and at most 2 approvals are outstanding), 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. |
rfcbot
added
proposed-final-comment-period
|
Note that AFAIK, this is still an issue: #27747 (comment) |
I'm confused by what you mean @BurntSushi. Isn't that impl still an orphan rules violation as pointed out in #27747 (comment)? What's the issue and how does this impact the FCP? |
|
@yaahc Ah sorry, yeah, I forgot about that. Yeah, regrettably, I agree to close this. It would be nice to improve things here, but it looks like a bigger issue than I thought and isn't going to be solved by this trait alone. |
|
🔔 This is now entering its final comment period, as per the review above. 🔔 |
rfcbot
added
final-comment-period
rfcbot
added
finished-final-comment-period
|
The final comment period, with a disposition to close, as per the review above, is now complete. As the automated representative of the governance process, I would like to thank the author for their work and everyone else who contributed. |
|
This issue is still referenced in at least two Line 78 in f0c4da4 It seems like those need to be rewritten, right? |
|
Is there a need for this trait to use |
|
That does sound like a mistake, but changing it now sounds impossible to do without breaking code. |
3 tasks
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
and others
The SliceConcatExt trait offers methods
concatandjoinon slices. For somewhat technical reasons, it wasn't possible to make these inherent methods.The methods themselves are stable, but the trait isn't. However, since the trait is in the prelude, the methods are usable in stable today.
Ideally, the methods would also be available on iterators, but there are performance ramifications for doing so. Impl specialization may help.
The text was updated successfully, but these errors were encountered: