Proposal for unifying traits and interfaces

Unifying traits and interfaces

There are three parts to this proposal:

• Adding default impls to ifaces
• Allowing iface composability
• Instance coherence: only one impl per iface/type pair

Then, rename iface to trait and that's it!

Adding default impls to ifaces

An example

In the middle::typeck::infer module (henceforth infer), there's a combine interface, and implementations of that interface for the three "type combiners" lub, sub, and glb. All three impls are required to implement all of the methods in the combine interface, even though some of the implementations are identical in two (or in all three!) of the type combiners. Right now, infer deals with this by defining an out-of-line method for each method for which there are multiple identical implementations, and having all the different implementations call the out-of-line-method.

For example, here's what it looks like for the modes method. In fact, there are nine methods in infer that are this way -- modes is just a representative example. The infcx method is also identical in all three, but since all it does is return *self, it's just identically implemented in all three.

iface combine {
    fn infcx() -> infer_ctxt;
    ...
    fn modes(a: ast::mode, b: ast::mode) -> cres<ast::mode>;
    ...
}

impl of combine for sub {
    fn infcx() -> infer_ctxt { *self }
    ...
    fn modes(a: ast::mode, b: ast::mode) -> cres<ast::mode> {
        super_modes(self, a, b)
    }
    ...
}

impl of combine for sub {
    fn infcx() -> infer_ctxt { *self }
    ...
    fn modes(a: ast::mode, b: ast::mode) -> cres<ast::mode> {
        super_modes(self, a, b)
    }
    ...
}

impl of combine for glb {
    fn infcx() -> infer_ctxt { *self }
    ...
    fn modes(a: ast::mode, b: ast::mode) -> cres<ast::mode> {
        super_modes(self, a, b)
    }
    ...
}

// Out-of-line method
fn super_modes<C:combine>(
    self: C, a: ast::mode, b: ast::mode)
    -> cres<ast::mode> {

    let tcx = self.infcx().tcx;
    ty::unify_mode(tcx, a, b)
}

Under this proposal, we could put the default implementation in the interface, so, instead of all of the above, we could just write:

trait combine {
    fn infcx() -> infer_ctxt { *self }
    ...
    fn modes(a: ast::mode, b: ast::mode) -> cres<ast::mode> {
        let tcx = self.infcx().tcx;
    	ty::unify_mode(tcx, a, b)
    }
    ...
}

impl of combine for sub {
    ... // only methods for which the default impl isn't enough
}

impl of combine for sub {
    ... // only methods for which the default impl isn't enough
}

impl of combine for glb {
    ... // only methods for which the default impl isn't enough
}

The only other thing that changed in this code was that the keyword iface changed to trait.

Allowing iface composability

(Note: the design in this section is broken. Fix forthcoming.)

Traits, as they appear in the literature, have a set of provided methods, implementing the behavior that a trait provides, and a (possibly empty) set of required methods that the provided methods can be written in terms of. For the required methods, only the names and types are specified, not the implementation. That suggests that in Rust, a trait's set of required methods could be specified using an iface. But if traits themselves are ifaces, then that means that ifaces can require ifaces. This goes along with the idea that traits/ifaces should be composable and order-independent: a trait C can extend traits A and B (in either order).

In today's Rust, a class or a type can implement interfaces A and B (in either order). For instance, in today's Rust you can write:

iface add { fn plus(n: int) -> int; }
iface subtract { fn minus(n: int) -> int; }

impl of add for int {
    fn plus(n: int) -> int { self + n }
}

impl of subtract for int {
    fn minus(n: int) -> int { self - n }
}

// int implements both add and subtract; order doesn't matter
fn main() { assert 3.plus(1) == 5.minus(1); }

But interfaces themselves aren't composable: we can't currently define an interface as the composition of more than one interface. Under this proposal, it would be possible to write:

iface arithmetic: add, subtract {
    ... // more methods here, if we want
}

The add, subtract part of the signature amounts to a set of required methods, in terms of which additional methods can be written.

Then, adding the ability to put default impls in ifaces (and changing the keyword from iface to trait), we get:

trait add { 
    fn plus(n: int) -> int { self + n }
}

trait subtract {
    fn minus(n: int) -> int { self - n }
}

trait arithmetic: add, subtract {
    ... // more methods here, if we want
}

impl int: arithmetic; // or something like this

fn main() { assert 3.plus(1) == 5.minus(1); }

The impl int: arithmetic; line is what associates the int type with the arithmetic trait. Syntax for this is still up in the air -- we could say int impls arithmetic; instead, or something else. (impl in this case is short for "implements", not short for "implementation", since now all implementation is happening in the traits!)

One place that could benefit from this so-called 'interface inheritance' is called out by a FIXME for issue #2616 in core::num. Although I have to think about it some more, I think we could clean up duplicated code between core/int_template.rs and core/uint_template.rs with this kind of strategy.

Conflict resolution

Traditional traits do some cool conflict resolution stuff when the traits being combined have methods with the same name, and we might want to do that eventually, but we can punt for now and just do what Rust already does if a type implements multiple interfaces that define a method with the same name, that is, raise a compile-time "multiple applicable methods in scope" error.

Instance coherence

(Note: the design in this section is broken. Fix forthcoming.)

In Rust, if you have an iface presenting a group of functions and not mandating any particular implementation -- as ifaces in Rust currently do -- then you can end up with the problem of different conflicting implementations for a particular type. This is known as the "instance coherence" problem (although in Rust perhaps we should call it the "implementation coherence"), or just the "coherence" problem for short.

Consider the following program, which compiles and runs in Rust today:

use std;

mod ht {
    iface hash {
        fn hash() -> uint;
        fn tostr() -> str;  // putting this into the interface is just
                            // a hack to give us a way to print
                            // keys. This doesn't go here at all.
    }

    type t<K,V> = [(K, V)];  // doesn't matter, we don't use it

    fn create<K:hash,V>() -> @t<K,V> {
        @[]/~
    }

    fn put<K:hash,V>(ht: @t<K,V>, k: K, v: V) {
        io::println(#fmt("ht put: %s hash to %ud", k.tostr(), k.hash()));
    }
}

mod Module1 {
    impl of ht::hash for uint {
        fn hash() -> uint { ret self; }
        fn tostr() -> str { ret #fmt("%ud", self); }  
    }
    fn foo() {
        let h = ht::create::<uint, str>();
        ht::put(h, 3u, "hi"); // 3u.hash() == 3u here
        Module2::bar(h);
    }
}

mod Module2 {
    impl of ht::hash for uint {
        fn hash() -> uint { ret self / 2; }
        fn tostr() -> str { ret #fmt("%ud", self); }
    }
    fn bar(h: @ht::t<uint, str>) {
        ht::put(h, 3u, "ho"); // 3u.hash() == 1u here
    }
}

fn main() {
    Module1::foo();
}

The output of this program is:

ht put: 3d hash to 3d
ht put: 3d hash to 1d

If put had really been inserting into a hash table instead of just printing, the table would end up with both "hi" and "ho" in it, even though we thought we were storing them under the same key, and "ho" should have overwritten "hi". The problem arises because there are conflicting implementations of the hash iface in Module1 and Module2. Neither one is wrong by itself, but when compiled together we get unexpected behavior.

To prevent this situation, we could do one of the following:

• Forbid method implementations anywhere except in traits (and in classes, maybe). The iface in the ht module would become a trait. The impl language form would become merely a way to associate types or classes with traits. We would add a line to each module like impl uint: ht::hash;.

• Keep impl around in its current form, but only allow an impl of a type for a particular trait to appear in the same crate as that trait.

To me, the first option is better, because the language is simpler if there are only one or two places where method implementations can appear, intead of two or three places.