Tracking issue for Minimal `impl Trait` (RFC 1522) #34511

Open
aturon opened this Issue Jun 27, 2016 · 32 comments

Projects

None yet
@aturon
Contributor
aturon commented Jun 27, 2016 edited

RFC 1522

Prior to stabilization, the following questions must be resolved:

  • Final syntax: whether impl Trait or some other syntax.
  • Whether to permit the feature in argument position as well.

These two questions are tied together, since some syntactic choices make sense only if argument position is allowed. A follow-up RFC is needed to resolve these questions differently than the initial (minimalistic) RFC.

Implementation has raised a number of interesting questions as well:

  • What is the precedence of the impl keyword when parsing types? Discussion: 1
  • Do we have to impose a DAG across all functions to allow for auto-safe leakage, or can we use some kind of deferral. Discussion: 1
  • How should we integrate impl trait into regionck? Discussion: 1, 2
@bstrie bstrie referenced this issue in rust-lang/rfcs Jun 28, 2016
Merged

Minimal `impl Trait` #1522

@pthariensflame
Contributor
pthariensflame commented Jul 15, 2016 edited

@aturon Can we actually put the RFC in the repository? (@mbrubeck commented there that this was a problem.)

@aturon
Contributor
aturon commented Jul 15, 2016

Done.

@eddyb
Member
eddyb commented Jul 31, 2016 edited

First attempt at implementation is #35091 (second, if you count my branch from last year).

One problem I ran into is with lifetimes. Type inference likes to put region variables everywhere and without any region-checking changes, those variables don't infer to anything other than local scopes.
However, the concrete type must be exportable, so I restricted it to 'static and explicitly named early-bound lifetime parameters, but it's never any of those if any function is involved - even a string literal doesn't infer to 'static, it's pretty much completely useless.

One thing I thought of, that would have 0 impact on region-checking itself, is to erase lifetimes:

  • nothing exposing the concrete type of an impl Trait should care about lifetimes - a quick search for Reveal::All suggests that's already the case in the compiler
  • a bound needs to be placed on all concrete types of impl Trait in the return type of a function, that it outlives the call of that function - this means that any lifetime is, by necessity, either 'static or one of the lifetime parameters of the function - even if we can't know which (e.g. "shortest of 'a and 'b")
  • we must choose a variance for the implicit lifetime parametrism of impl Trait (i.e. on all lifetime parameters in scope, same as with type parameters): invariance is easiest and gives more control to the callee, while contravariance lets the caller do more and would require checking that every lifetime in the return type is in a contravariant position (same with covariant type parametrism instead of invariant)
  • the auto trait leaking mechanism requires that a trait bound may be put on the concrete type, in another function - since we've erased the lifetimes and have no idea what lifetime goes where, every erased lifetime in the concrete type will have to be substituted with a fresh inference variable that is guaranteed to not be shorter than the shortest lifetime out of all actual lifetime parameters; the problem lies in the fact that trait impls can end up requiring stronger lifetime relationships (e.g. X<'a, 'a> or X<'static>), which must be detected and errored on, as they can't be proven for those lifetimes

That last point about auto trait leakage is my only worry, everything else seems straight-forward.
It's not entirely clear at this point how much of region-checking we can reuse as-is. Hopefully all.

cc @rust-lang/lang

@arielb1
Contributor
arielb1 commented Jul 31, 2016 edited

@eddyb

But lifetimes are important with impl Trait - e.g.

fn get_debug_str(s: &str) -> impl fmt::Debug {
    s
}

fn get_debug_string(s: &str) -> impl fmt::Debug {
    s.to_string()
}

fn good(s: &str) -> Box<fmt::Debug+'static> {
    // if this does not compile, that would be quite annoying
    Box::new(get_debug_string())
}

fn bad(s: &str) -> Box<fmt::Debug+'static> {
    // if this *does* compile, we have a problem
    Box::new(get_debug_str())
}

I mentioned that several times in the RFC threads

@arielb1
Contributor
arielb1 commented Jul 31, 2016 edited

trait-object-less version:

fn as_debug(s: &str) -> impl fmt::Debug;

fn example() {
    let mut s = String::new("hello");
    let debug = as_debug(&s);
    s.truncate(0);
    println!("{:?}", debug);
}

This is either UB or not depending on the definition of as_debug.

@eddyb
Member
eddyb commented Jul 31, 2016

@arielb1 Ah, right, I forgot that one of the reasons I did what I did was to only capture lifetime parameters, not anonymous late-bound ones, except it doesn't really work.

@eddyb
Member
eddyb commented Jul 31, 2016

@arielb1 Do we have a strict outlives relation we can put between lifetimes found in the concrete type pre-erasure and late-bound lifetimes in the signature? Otherwise, it might not be a bad idea to just look at lifetime relationships and insta-fail any direct or indirect 'a outlives 'b where 'a is anything other than 'static or a lifetime parameter and 'b appears in the concrete type of an impl Trait.

@nikomatsakis
Contributor

Sorry for taking a while to write back here. So I've been thinking
about this problem. My feeling is that we do, ultimately, have to (and
want to) extend regionck with a new kind of constraint -- I'll call it
an \in constraint, because it allows you to say something like '0 \in {'a, 'b, 'c}, meaning that the region used for '0 must be
either 'a, 'b, or 'c. I'm not sure of the best way to integrate
this into solving itself -- certainly if the \in set is a singleton
set, it's just an equate relation (which we don't currently have as a
first-class thing, but which can be composed out of two bounds), but
otherwise it makes things complicated.

This all relates to my desire to make the set of region constraints
more expressive than what we have today. Certainly one could compose a
\in constraint out of OR and == constraints. But of course more
expressive constraints are harder to solve and \in is no different.

Anyway, let me just lay out a bit of my thinking here. Let's work with this
example:

pub fn foo<'a,'b>(x: &'a [u32], y: &'b [u32]) -> impl Iterator<Item=u32> {...}

I think the most accurate desugaring for a impl Trait is probably a
new type:

pub struct FooReturn<'a, 'b> {
    field: XXX // for some suitable type XXX
}

impl<'a,'b> Iterator for FooReturn<'a,'b> {
    type Item = <XXX as Iterator>::Item;
}

Now the impl Iterator<Item=u32> in foo should behave the same as
FooReturn<'a,'b> would behave. It's not a perfect match though. One
difference, for example, is variance, as eddyb brought up -- I am
assuming we will make impl Foo-like types invariant over the type
parameters of foo. The auto trait behavior works out, however.
(Another area where the match might not be ideal is if we ever add the
ability to "pierce" the impl Iterator abstraction, so that code
"inside" the abstraction knows the precise type -- then it would sort
of have an implicit "unwrap" operation taking place.)

In some ways a better match is to consider a kind of synthetic trait:

trait FooReturn<'a,'b> {
    type Type: Iterator<Item=u32>;
}

impl<'a,'b> FooReturn<'a,'b> for () {
    type Type = XXX;
}

Now we could consider the impl Iterator type to be like <() as FooReturn<'a,'b>>::Type. This is also not a perfect match, because we
would ordinarily normalize it away. You might imagine using specialization
to prevent that though:

trait FooReturn<'a,'b> {
    type Type: Iterator<Item=u32>;
}

impl<'a,'b> FooReturn<'a,'b> for () {
    default type Type = XXX; // can't really be specialized, but wev
}

In this case, <() as FooReturn<'a,'b>>::Type would not normalize,
and we have a much closer match. The variance, in particular, behaves
right; if we ever wanted to have some type that are "inside" the
abstraction, they would be the same but they are allowed to
normalize. However, there is a catch: the auto trait stuff doesn't
quite work. (We may want to consider harmonizing things here,
actually.)

Anyway, my point in exploring these potential desugarings is not to
suggest that we implement "impl Trait" as an actual desugaring
(though it might be nice...) but to give an intuition for our job. I
think that the second desugaring -- in terms of projections -- is a
pretty helpful one for guiding us forward.

One place that this projection desugaring is a really useful guide is
the "outlives" relation. If we wanted to check whether <() as FooReturn<'a,'b>>::Type: 'x, RFC 1214 tells us that we can prove this
so long as 'a: 'x and 'b: 'x holds. This is I think how we want
to handle things for impl trait as well.

At trans time, and for auto-traits, we will have to know what XXX
is, of course. The basic idea here, I assume, is to create a type
variable for XXX and check that the actual values which are returned
can all be unified with XXX. That type variable should, in theory,
tell us our answer. But of course the problem is that this type
variable may refer to a lot of regions which are not in scope in the
fn signature -- e.g., the regions of the fn body. (This same problem
does not occur with types; even though, technically, you could put
e.g. a struct declaration in the fn body and it would be unnameable,
that's a kind of artificial restriction -- one could just as well move
the struct outside the fn.)

If you look both at the struct desugaring or the impl, there is an
(implicit in the lexical structure of Rust) restriction that XXX can
only name either 'static or lifetimes like 'a and 'b, which
appear in the function signature. That is the thing we are not
modeling here. I'm not sure the best way to do it -- some type
inference schemes have a more direct representation of scoping, and
I've always wanted to add that to Rust, to help us with closures. But
let's think about smaller deltas first I guess.

This is where the \in constraint comes from. One can imagine adding
a type-check rule that (basically) FR(XXX) \subset {'a, 'b} --
meaning that the "free regions" appearing in XXX can only be 'a and
'b. This would wind up translating to \in requirements for the
various regions that appear in XXX.

Let's look at an actual example:

fn foo<'a,'b>(x: &'a [u32], y: &'b [u32]) -> impl Iterator<Item=u32> {
    if condition { x.iter().cloned() } else { y.iter().cloned() }
}

Here, the type if condition is true would be something like
Cloned<SliceIter<'a, i32>>. But if condition is false, we would
want Cloned<SliceIter<'b, i32>>. Of course in both cases we would
wind up with something like (using numbers for type/region variables):

Cloned<SliceIter<'0, i32>> <: 0
'a: '0 // because the source is x.iter()
Cloned<SliceIter<'1, i32>> <: 0
'b: '1 // because the source is y.iter()

If we then instantiate the variable 0 to Cloned<SliceIter<'2, i32>>,
we have '0: '2 and '1: '2, or a total set of region relations
like:

'a: '0
'0: '2
'b: '1
'1: '2
'2: 'body // the lifetime of the fn body

So what value should we use for '2? We have also the additional
constraint that '2 in {'a, 'b}. With the fn as written, I think we
would have to report an error, since neither 'a nor 'b is a
correct choice. Interestingly, though, if we added the constraint 'a: 'b, then there would be a correct value ('b).

Note that if we just run the normal algorithm, we would wind up with
'2 being 'body. I'm not sure how to handle the \in relations
except for exhaustive search (though I can imagine some special
cases).

OK, that's as far as I've gotten. =)

@nikomatsakis
Contributor

On the PR #35091, @arielb1 wrote:

I don't like the "capture all lifetimes in the impl trait" approach and would prefer something more like lifetime elision.

I thought it would make more sense to discuss here. @arielb1, can you elaborate more on what you have in mind? In terms of the analogies I made above, I guess you are fundamentally talking about "pruning" the set of lifetimes that would appear either as parameters on the newtype or in the projection (i.e., <() as FooReturn<'a>>::Type instead of <() as FooReturn<'a,'b>>::Type or something?

I don't think that the lifetime elision rules as they exist would be a good guide in this respect: if we just picked the lifetime of &self to include only, then we wouldn't necessarily be able to include the type parameters from the Self struct, nor type parameters from the method, since they may have WF conditions that require us to name some of the other lifetimes.

Anyway, it'd be great to see some examples that illustrate the rules you have in mind, and perhaps any advantages thereof. :) (Also, I guess we would need some syntax to override the choice.) All other things being equal, if we can avoid having to pick from N lifetimes, I'd prefer that.

@bors bors added a commit that referenced this issue Aug 12, 2016
@bors bors Auto merge of #35091 - eddyb:impl-trait, r=nikomatsakis
Implement `impl Trait` in return type position by anonymization.

This is the first step towards implementing `impl Trait` (cc #34511).
`impl Trait` types are only allowed in function and inherent method return types, and capture all named lifetime and type parameters, being invariant over them.
No lifetimes that are not explicitly named lifetime parameters are allowed to escape from the function body.
The exposed traits are only those listed explicitly, i.e. `Foo` and `Clone` in `impl Foo + Clone`, with the exception of "auto traits" (like `Send` or `Sync`) which "leak" the actual contents.

The implementation strategy is anonymization, i.e.:
```rust
fn foo<T>(xs: Vec<T>) -> impl Iterator<Item=impl FnOnce() -> T> {
    xs.into_iter().map(|x| || x)
}

// is represented as:
type A</*invariant over*/ T> where A<T>: Iterator<Item=B<T>>;
type B</*invariant over*/ T> where B<T>: FnOnce() -> T;
fn foo<T>(xs: Vec<T>) -> A<T> {
    xs.into_iter().map(|x| || x): $0 where $0: Iterator<Item=$1>, $1: FnOnce() -> T
}
```
`$0` and `$1` are resolved (to `iter::Map<vec::Iter<T>, closure>` and the closure, respectively) and assigned to `A` and `B`, after checking the body of `foo`. `A` and `B` are *never* resolved for user-facing type equality (typeck), but always for the low-level representation and specialization (trans).

The "auto traits" exception is implemented by collecting bounds like `impl Trait: Send` that have failed for the obscure `impl Trait` type (i.e. `A` or `B` above), pretending they succeeded within the function and trying them again after type-checking the whole crate, by replacing `impl Trait` with the real type.

While passing around values which have explicit lifetime parameters (of the function with `-> impl Trait`) in their type *should* work, regionck appears to assign inference variables in *way* too many cases, and never properly resolving them to either explicit lifetime parameters, or `'static`.
We might not be able to handle lifetime parameters in `impl Trait` without changes to lifetime inference, but type parameters can have arbitrary lifetimes in them from the caller, so most type-generic usecases (or not generic at all) should not run into this problem.

cc @rust-lang/lang
f55ac69
@petrochenkov
Contributor
petrochenkov commented Aug 12, 2016 edited

I haven't seen interactions of impl Trait with privacy discussed anywhere.
Now fn f() -> impl Trait can return a private type S: Trait similarly to trait objects fn f() -> Box<Trait>. I.e. objects of private types can walk freely outside of their module in anonymized form.
This seems reasonable and desirable - the type itself is an implementation detail, only its interface, available through a public trait Trait is public.
However there's one difference between trait objects and impl Trait. With trait objects alone all trait methods of private types can get internal linkage, they will still be callable through function pointers. With impl Traits trait methods of private types are directly callable from other translation units. The algorithm doing "internalization" of symbols will have to try harder to internalize methods only for types not anonymized with impl Trait, or to be very pessimistic.

@arielb1
Contributor
arielb1 commented Aug 14, 2016

@nikomatsakis

The "explicit" way to write foo would be

fn foo<'a: 'c,'b: 'c,'c>(x: &'a [u32], y: &'b [u32]) -> impl Iterator<Item=u32> + 'c {
    if condition { x.iter().cloned() } else { y.iter().cloned() }
}

Here there is no question about the lifetime bound. Obviously, having to write the lifetime bound each time would be quite repetitive. However, the way we deal with that kind of repetition is generally through lifetime elision. In the case of foo, elision would fail and force the programmer to explicitly specify lifetimes.

I am opposed to adding explicitness-sensitive lifetime elision as @eddyb did only in the specific case of impl Trait and not otherwise.

@jamesmunns jamesmunns referenced this issue in Covertness/coap-rs Aug 14, 2016
Open

static handler brings problems #10

@nikomatsakis
Contributor

@arielb1 hmm, I'm not 100% sure how to think about this proposed syntax in terms of the "desugarings" that I discussed. It allows you to specify what appears to be a lifetime bound, but the thing we are trying to infer is mostly what lifetimes appear in the hidden type. Does this suggest that at most one lifetime could be "hidden" (and that it would have to be specified exactly?)

It seems like it's not always the case that a "single lifetime parameter" suffices:

fn foo<'a, 'b>(x: &'a [u32], y: &'b [u32]) -> impl Iterator<Item=u32> {
    x.iter().chain(y).cloned()
}

In this case, the hidden iterator type refers to both 'a and 'b (although it is variant in both of them; but I guess we could come up with an example that is invariant).

@nikomatsakis
Contributor
nikomatsakis commented Aug 18, 2016 edited

So @aturon and I discussed this issue somewhat and I wanted to share. There are really a couple of orthogonal questions here and I want to separate them out. The first question is "what type/lifetime parameters can potentially be used in the hidden type?" In terms of the (quasi-)desugaring into a default type, this comes down to "what type parameters appear on the trait we introduce". So, for example, if this function:

fn foo<'a, 'b, T>() -> impl Trait { ... }

would get desugared to something like:

fn foo<'a, 'b, T>() -> <() as Foo<...>>::Type { ... }
trait Foo<...> {
  type Type: Trait;
}
impl<...> Foo<...> for () {
  default type Type = /* inferred */;
}

then this question comes down to "what type parameters appear on the trait Foo and its impl"? Basically, the ... here. Clearly this include include the set of type parameters that appear are used by Trait itself, but what additional type parameters? (As I noted before, this desugaring is 100% faithful except for the leakage of auto traits, and I would argue that we should leak auto traits also for specializable impls.)

The default answer we've been using is "all of them", so here ... would be 'a, 'b, T (along with any anonymous parameters that may appear). This may be a reasonable default, but it's not necessarily the best default. (As @arielb1 pointed out.)

This has an effect on the outlives relation, since, in order to determine that <() as Foo<...>>::Type (referring to some particular, opaque instantiation of impl Trait) outlives 'x, we effectively must show that ...: 'x (that is, every lifetime and type parameter).

This is why I say it is not enough to consider lifetime parameters: imagine that we have some call to foo like foo::<'a0, 'b0, &'c0 i32>. This implies that all three lifetimes, '[abc]0, must outlive 'x -- in other words, so long as the return value is in use, this will prolog the loans of all data given into the function. But, as @arielb1 poitned out, elision suggests that this will usually be longer than necessary.

So I imagine that what we need is:

  • to settle on a reasonable default, perhaps using intution from elision;
  • to have an explicit syntax for when the default is not appropriate.

@aturon spitballed something like impl<...> Trait as the explicit syntax, which seems reasonable. Therefore, one could write:

fn foo<'a, 'b, T>(...) -> impl<T> Trait { }

to indicate that the hidden type does not in fact refer to 'a or 'b but only T. Or one might write impl<'a> Trait to indicate that neither 'b nor T are captured.

As for the defaults, it seems like having more data would be pretty useful -- but the general logic of elision suggests that we would do well to capture all the parameters named in the type of self, when applicable. E.g., if you have fn foo<'a,'b>(&'a self, v: &'b [u8]) and the type is Bar<'c, X>, then the type of self would be &'a Bar<'c, X> and hence we would capture 'a, 'c, and X by default, but not 'b.


Another related note is what the meaning of a lifetime bound is. I think that sound lifetime bounds have an existing meaning that should not be changed: if we write impl (Trait+'a) that means that the hidden type T outlives 'a. Similarly one can write impl (Trait+'static) to indicate that there are no borrowed pointers present (even if some lifetimes are captured). When inferring the hidden type T, this would imply a lifetime bound like $T: 'static, where $T is the inference variable we create for the hidden type. This would be handled in the usual way. From a caller's perspective, where the hidden type is, well, hidden, the 'static bound would allow us to conclude that impl (Trait+'static) outlives 'static even if there are lifetime parameters captured.

Here it just behaves exactly as the desugaring would behave:

fn foo<'a, 'b, T>() -> <() as Foo<'a, 'b, 'T>>::Type { ... }
trait Foo<'a, 'b, T> {
  type Type: Trait + 'static; // <-- note the `'static` bound appears here
}
impl<'a, 'b, T> Foo<...> for () {
  default type Type = /* something that doesn't reference `'a`, `'b`, or `T` */;
}

All of this is orthogonal from inference. We still want (I think) to add the notion of a "choose from" constraint and modify inference with some heuristics and, possibly, exhaustive search (the experience from RFC 1214 suggests that heuristics with a conservative fallback can actually get us very far; I'm not aware of people running into limitations in this respect, though there is probably an issue somewhere). Certainly, adding lifetime bounds like 'static or 'a` may influence inference, and thus be helpful, but that is not a perfect solution: for one thing, they are visible to the caller and become part of the API, which may not be desired.

@arielb1
Contributor
arielb1 commented Aug 18, 2016

Possible options:

Explicit lifetime bound with output parameter elision

Like trait objects today, impl Trait objects have a single lifetime bound parameter, which is inferred using the elision rules.

Disadvantage: unergonomic
Advantage: clear

Explicit lifetime bounds with "all generic" elision

Like trait objects today, impl Trait objects have a single lifetime bound parameter.

However, elision creates a new early-bound parameters that outlives all explicit parameters:

fn foo<T>(&T) -> impl Foo
-->
fn foo<'total, T: 'total>(&T) -> impl Foo + 'total

Disadvantage: adds an early-bound parameter

more.

@nrc nrc added the T-lang label Aug 19, 2016
@Boscop
Boscop commented Nov 15, 2016

I ran into this issue with impl Trait +'a and borrowing: #37790

@IanWhitney

If I'm understanding this change correctly (and the chance of that is probably low!), then I think this playground code should work:

https://play.rust-lang.org/?gist=496ec05e6fa9d3a761df09c95297aa2a&version=nightly&backtrace=0

Both ThingOne and ThingTwo implement the Thing trait. build says it will return something that implements Thing, which it does. Yet it does not compile. So I'm clearly misunderstanding something.

@eddyb
Member
eddyb commented Jan 9, 2017

That "something" must have a type, but in your case you have two conflicting types. @nikomatsakis has previously suggested making this work in general by creating e.g. ThingOne | ThingTwo as type mismatches appear.

@WiSaGaN
Contributor
WiSaGaN commented Jan 9, 2017 edited

@eddyb could you elaborate on ThingOne | ThingTwo? Don't you need to have Box if we only know the type at run-time? Or is it a kind of enum?

@eddyb
Member
eddyb commented Jan 9, 2017

Yeah it could be an ad-hoc enum-like type that delegated trait method calls, where possible, to its variants.

@glaebhoerl
Contributor

I've wanted that kind of thing before too. The anonymous enums RFC: rust-lang/rfcs#1154

@eddyb
Member
eddyb commented Jan 9, 2017

It's a rare case of something that works better if it's inference-driven, because if you only create these types on a mismatch, the variants are different (which is a problem with the generalized form).
Also you can get something out of not having pattern-matching (except in obviously disjoint cases?).
But IMO delegation sugar would "just work" in all relevant cases, even if you manage to get a T | T.

@glaebhoerl
Contributor

Could you spell out the other, implicit halves of those sentences? I don't understand most of it, and suspect I'm missing some context. Were you implicitly responding to the problems with union types? That RFC is simply anonymous enums, not union types - (T|T) would be exactly as problematic as Result<T, T>.

@eddyb
Member
eddyb commented Jan 9, 2017

Oh, nevermind, I got the proposals confused (also stuck on mobile until I sort out my failing HDD so apologies for sounding like on Twitter).

I find (positional, i.e T|U != U|T) anonymous enums intriguing, and I believe they could be experimented with in a library if we had variadic generics (you can side-step this by using hlist) and const generics (ditto, with peano numbers).

But, at the same time, if we had language support for something, it'd be union types, not anonymous enums. E.g. not Result but error types (to bypass the tedium of named wrappers for them).

@kud1ing
Contributor
kud1ing commented Jan 16, 2017 edited

I am not sure whether this is the righ place to ask, but why is a keyword like impl needed? I could not find a discussion (could be my fault).

If a function returns impl Trait, its body can return values of any type that implements Trait

Since

fn bar(a: &Foo) {
  ...
}

means "accept a reference to a type that implements trait Foo" i would expect

fn bar() -> Foo {
  ...
}

to mean "return a type that implements trait Foo". Is this impossible?

@Nemo157
Nemo157 commented Jan 16, 2017 edited

@kud1ing the reason is to not remove the possibility of having a function that returns the dynamically sized type Trait if support for dynamically sized return values is added in the future. Currently Trait is already a valid DST, it's just not possible to return a DST so you need to box it to make it a sized type.

EDIT: There is some discussion about this on the linked RFC thread.

@Enet4
Enet4 commented Jan 16, 2017 edited

Well, for one, regardless of whether dynamically sized return values will be added, I prefer the current syntax. Unlike what happens with trait objects, this isn't type erasure, and any coincidences like "parameter f: &Foo takes something that impls Foo, whereas this returns something that impls Foo" could be misleading.

@Nercury
Contributor
Nercury commented Jan 16, 2017

I gathered from RFC discussion that right now impl is a placeholder implementation, and no impl is very much desired. Is there any reason for not wanting an impl Trait if the return value is not DST?

@nikomatsakis
Contributor

I think the current impl technique for handling "auto trait leakage" is problematic. We should instead enforce a DAG ordering so that if you define a fn fn foo() -> impl Iterator, and you have a caller fn bar() { ... foo() ... }, then we have to type-check foo() before bar() (so that we know what the hidden type is). If a cycle results, we'd report an error. This is a conservative stance -- we can probably do better -- but I think the current technique, where we collect auto-trait obligations and check them at the end, does not work in general. For example, it would not work well with specialization.

(Another possibility that might be more permissive than requiring a strict DAG is to type-check both fns "together" to some extent. I think that is something to consider only after we have re-archicted the trait system impl a bit.)

@nikomatsakis
Contributor

@Nercury I don't understand. Are you asking if there are reasons to not want fn foo() -> Trait to mean -> impl Trait?

@Nercury
Contributor
Nercury commented Jan 17, 2017 edited

@nikomatsakis Yes, I was asking precisely that, sorry for convulted language :). I thought that doing this without impl keyword would be simpler, because this behavior is exactly what one would expect (when a concrete type is returned in place of trait return type). However, I might be missing something, that's why I was asking.

@rpjohnst
Contributor

The difference is that functions returning impl Trait always return the same type- it's basically return type inference. IIUC, functions returning just Trait would be able to return any implementation of that trait dynamically, but the caller would need to be prepared to allocate space for the return value via something like box foo().

@Ixrec
Ixrec commented Jan 17, 2017

@Nercury The simple reason is that the -> Trait syntax already has a meaning, so we have to use something else for this feature.

I've actually seen people expect both kinds of behavior by default, and this sort of confusion comes up often enough I'd honestly rather that fn foo() -> Trait not mean anything (or be a warning by default) and there were explicit keywords for both the "some type known at compile time that I get to choose but the caller doesn't see" case and the "trait object that could be dynamically dispatching to any type implementing Trait" case, e.g. fn foo() -> impl Trait vs fn foo() -> dyn Trait. But obviously those ships have sailed.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment