Typeclass/interface implementation #1227

jwise · 2011-11-30T00:00:37Z

This is a stub bug, tracking @marijnh's efforts in implementing typeclasses. Created so that I could link to it from another bug.

More detail: https://mail.mozilla.org/pipermail/rust-dev/2011-November/001010.html

jwise · 2011-11-30T00:01:08Z

Oops, tagged the wrong Marijn. Now tagging @marijnh.

marijnh · 2011-11-30T11:00:36Z

Latest proposal at https://github.com/graydon/rust/wiki/Interface-and-Implementation-Proposal

boggle · 2011-12-01T22:18:25Z

I like the direction this is taking, especially the unification of static type classes and the proposal for boxed existentials (interface, value).

Have added an idea on how to add delegation to the wiki.

Issue #1227

No conflict resolution or polymorphism yet. Issue #1227

Get rid of expr_self_call, introduces def_self. `self` is now, syntactically, simply a variable. A method implicitly brings a `self` binding into scope. Issue #1227

Issue #1227

Something will still have to be done to the AST to make it possible to say `x.foo::<int>()`, since currently field access never allows type parameters. Issue #1227

(No one is actually reading it yet.) Issue #1227

Issue #1227

And fix a bug where importing anything from an external module would import all impls in the module. Issue #1227

We should probalby warn when defining a method foo on {foo: int} etc. This should reduce the amount of useless typevars that are allocated. Issue #1227

Issue #1227

This way, you can explicitly provide type parameters when calling a generic method. Issue #1227

The parameters of the impl weren't being combined in the right way with the parameters of the methods. The test worked only by accident. Issue #1227

Issue #1227

Extern interfaces still don't get recognized. Issue #1227

Issue #1227

kud1ing · 2012-01-03T10:12:19Z

I am not sure, does the current approach/plan include default implementations?
For example the interface

iface seq<T> {
    fn len() -> uint;
    fn iter(block(T));
}

could implement a function any() like vec::any() and it would be available to all sequences like list, string etc.

marijnh · 2012-01-03T14:33:59Z

Not in the first iteration. They're probably useful enough to be included at some later point.

Reached a point where simple uses of interfaces without bounds work. Issue #1227

The trick of interpreting parameters to the iface type as parameters to the impl was just too magical. Issue #1227

Issue #1227

marijnh · 2012-01-03T15:39:47Z

New demo of what's possible: https://github.com/graydon/rust/blob/master/src/test/run-pass/iface-to-str.rs

kud1ing · 2012-01-03T16:25:46Z

Very nice.
I catch myself thinking "this should be implemented using type classes / interfaces" very often recently.

kud1ing · 2012-01-03T16:40:52Z

I wonder whether we should drop "of":

impl of to_str for int

vs

impl to_str for int

The latter contains less words and characters to mentally process, but has no diminished clearness to me.
Or is necessary syntax-wise?

marijnh · 2012-01-03T16:54:29Z

We can't drop of, it is needed to allow forms where a specific name for the impl is given, and for impls without an explicit interface.

Aligning the type parameters of the ifaces, impls, and methods correctly in typeck is almost brain surgery. Seems to work now for everything I threw at it, but might still break in other corner cases. Issue #1227

I.e. fn foo<T: seq<T>>(...). This leads to weird circularities that seem to never make any sense, so it seems prudent to forbid it. Issue #1227

Issue #1227

Also, move checking of ifaces into the collect phase of typeck to give further passes some guarantees. Issue #1227

Issue #1227

marijnh · 2012-01-05T13:35:18Z

The basics are working now. I've split off the remaining work into #1436 #1437 #1438 and #1439 . Closing this.

kud1ing · 2012-01-06T10:28:48Z

Have you found that naming implementations is strictly necessary?
For importing/exporting i'd think we could just import/export the implemented functions, can't we?

Only enable JIT tests on x86_64

* Improve mem::swap test to cover more types * Remove the mem::swap() hook

jwise mentioned this issue Nov 30, 2011

Native pointer vectors #1167

Closed

kud1ing mentioned this issue Dec 13, 2011

Implement complex numbers #1284

Closed

ghost assigned marijnh Dec 15, 2011

marijnh added a commit that referenced this issue Dec 16, 2011

Parse and resolve implementations.

888bc80

Issue #1227

marijnh added a commit that referenced this issue Dec 16, 2011

Get a very primitive form of typechecking/resolving to work for impls

6a16f57

No conflict resolution or polymorphism yet. Issue #1227

marijnh added a commit that referenced this issue Dec 16, 2011

Make uses of self in impls compile

7efef98

Get rid of expr_self_call, introduces def_self. `self` is now, syntactically, simply a variable. A method implicitly brings a `self` binding into scope. Issue #1227

marijnh added a commit that referenced this issue Dec 16, 2011

Improve resolution of impls

14fcb13

Issue #1227

marijnh added a commit that referenced this issue Dec 16, 2011

Make polymorphic impl methods work

d529757

Something will still have to be done to the AST to make it possible to say `x.foo::<int>()`, since currently field access never allows type parameters. Issue #1227

marijnh added a commit that referenced this issue Dec 16, 2011

Write impl data to crate library files

0a36261

(No one is actually reading it yet.) Issue #1227

marijnh added a commit that referenced this issue Dec 16, 2011

Finish resolving and calling of crate-external impls

58a81a6

Issue #1227

marijnh added a commit that referenced this issue Dec 18, 2011

Improve efficienty of impl cache in resolve

b36ade1

And fix a bug where importing anything from an external module would import all impls in the module. Issue #1227

marijnh added a commit that referenced this issue Dec 18, 2011

Only look for a matching method when normal field access fails

f4acaf6

We should probalby warn when defining a method foo on {foo: int} etc. This should reduce the amount of useless typevars that are allocated. Issue #1227

marijnh added a commit that referenced this issue Dec 19, 2011

Complain when multiple methods in the same scope match a call

4d524b9

Issue #1227

marijnh added a commit that referenced this issue Dec 19, 2011

Add type argument field to expr_path

60acae4

This way, you can explicitly provide type parameters when calling a generic method. Issue #1227

marijnh added a commit that referenced this issue Dec 19, 2011

Fix bug in type parameter handling for impl methods

e4e2d6d

The parameters of the impl weren't being combined in the right way with the parameters of the methods. The test worked only by accident. Issue #1227

marijnh added a commit that referenced this issue Dec 28, 2011

Change representation of type params to handle interface bounds

1ed6a27

Issue #1227

marijnh added a commit that referenced this issue Jan 2, 2012

Check that type parameter bounds are interface types

40d5f28

Issue #1227

marijnh added a commit that referenced this issue Jan 2, 2012

More resolving and typechecking of bounded type parameters.

664a044

Extern interfaces still don't get recognized. Issue #1227

marijnh added a commit that referenced this issue Jan 2, 2012

Write out vtables for interface implementations

1741ef7

Issue #1227

marijnh added a commit that referenced this issue Jan 3, 2012

More work on translating dictionary-passing

e1dc40b

Reached a point where simple uses of interfaces without bounds work. Issue #1227

marijnh added a commit that referenced this issue Jan 3, 2012

Make syntax for impls less magical

506a6ec

The trick of interpreting parameters to the iface type as parameters to the impl was just too magical. Issue #1227

marijnh added a commit that referenced this issue Jan 3, 2012

Wire in resolution of param bounds for method calls

5ea3c96

Issue #1227

marijnh added a commit that referenced this issue Jan 3, 2012

Make resolution of dictionaries on bounded params work

4e88d5a

Issue #1227

marijnh added a commit that referenced this issue Jan 4, 2012

Prevent typenames in param bounds from resolving to their own param

94d40be

I.e. fn foo<T: seq<T>>(...). This leads to weird circularities that seem to never make any sense, so it seems prudent to forbid it. Issue #1227

marijnh added a commit that referenced this issue Jan 4, 2012

Add test for generic iface methods

b02a77d

Issue #1227

marijnh added a commit that referenced this issue Jan 5, 2012

Write the iface type of an impl in the crate data

6559aa8

Also, move checking of ifaces into the collect phase of typeck to give further passes some guarantees. Issue #1227

marijnh added a commit that referenced this issue Jan 5, 2012

Get external interfaces / impls working

3a20dda

Issue #1227

marijnh closed this as completed Jan 5, 2012

jwise unassigned marijnh Jun 16, 2014

bjorn3 added a commit to bjorn3/rust that referenced this issue May 15, 2022

Merge pull request rust-lang#1227 from Kmeakin/master

533f9d8

Only enable JIT tests on x86_64

celinval pushed a commit to celinval/rust-dev that referenced this issue Jun 4, 2024

Remove the mem::swap() hook (rust-lang#1227)

ed566c1

* Improve mem::swap test to cover more types * Remove the mem::swap() hook

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Typeclass/interface implementation #1227

Typeclass/interface implementation #1227

jwise commented Nov 30, 2011

jwise commented Nov 30, 2011

marijnh commented Nov 30, 2011

boggle commented Dec 1, 2011

kud1ing commented Jan 3, 2012

marijnh commented Jan 3, 2012

marijnh commented Jan 3, 2012

kud1ing commented Jan 3, 2012

kud1ing commented Jan 3, 2012

marijnh commented Jan 3, 2012

marijnh commented Jan 5, 2012

kud1ing commented Jan 6, 2012

Typeclass/interface implementation #1227

Typeclass/interface implementation #1227

Comments

jwise commented Nov 30, 2011

jwise commented Nov 30, 2011

marijnh commented Nov 30, 2011

boggle commented Dec 1, 2011

kud1ing commented Jan 3, 2012

marijnh commented Jan 3, 2012

marijnh commented Jan 3, 2012

kud1ing commented Jan 3, 2012

kud1ing commented Jan 3, 2012

marijnh commented Jan 3, 2012

marijnh commented Jan 5, 2012

kud1ing commented Jan 6, 2012