A proc macro to ease development using Inversion of Control patterns in Rust.
entrait
is used to generate a trait from the definition of a regular function.
The main use case for this is that other functions may depend upon the trait
instead of the concrete implementation, enabling better test isolation.
The macro looks like this:
```rust
fn my_function
which generates the trait MyFunction
:
rust
trait MyFunction {
fn my_function(&self);
}
my_function
's first and only parameter is deps
which is generic over some unknown type D
.
This would correspond to the self
parameter in the trait.
But what is this type supposed to be? We can generate an implementation in the same go, using for Type
:
```rust struct App;
fn my_function
// Generated: // trait MyFunction { // fn myfunction(&self); // } // // impl MyFunction for App { // fn myfunction(&self) { // my_function(self) // } // }
fn main() { let app = App; app.my_function(); } ```
The advantage of this pattern comes into play when a function declares its dependencies, as trait bounds:
```rust
fn foo(deps: &(impl Bar)) { deps.bar(); }
fn bar
The functions may take any number of parameters, but the first one is always considered specially as the "dependency parameter".
Functions may also be non-generic, depending directly on the App:
```rust
fn extractsomething(app: &App) -> SomeType { app.something } ```
These kinds of functions may be considered "leaves" of a dependency tree.
The idea behind entrait
is to explore a specific architectural pattern:
* Interfaces with one runtime implementation
* named traits as the interface of single functions
entrait
does not implement Dependency Injection (DI). DI is a strictly object-oriented concept that will often look awkward in Rust.
The author thinks of DI as the "reification of code modules": In a DI-enabled programming environment, code modules are grouped together
as objects and other modules may depend upon the interface of such an object by receiving some instance that implements it.
When this pattern is applied successively, one ends up with an in-memory dependency graph of high-level modules.
entrait
tries to turn this around by saying that the primary abstraction that is depended upon is a set of functions, not a set of code modules.
An architectural consequence is that one ends up with one ubiquitous type that represents a running application that implements all these function abstraction traits. But the point is that this is all loosely coupled: Most function definitions themselves do not refer to this god-like type, they only depend upon traits.
async
supportSince Rust at the time of writing does not natively support async methods in traits, you may opt in to having #[async_trait]
generated
for your trait:
```rust
async fn foo
This is designed to be forwards compatible with real async fn in traits. When that day comes, you should be able to just remove the
async_trait=true`
to get a proper zero-cost future.
The macro supports autogenerating [mockall] mock structs:
```rust
fn foo
fn my_func(deps: &(impl Foo)) -> u32 { deps.foo() }
fn main() {
let mut deps = MockFoo::new();
deps.expectfoo().returning(|| 42);
asserteq!(42, my_func(&deps));
}
``
Using
mockallis easy enough when there is only one trait bound, because the generated trait need only be attributed with
mockall::automock`.
unimock
With multiple trait bounds, this becomes a little harder: We need some concrete struct that implement all the given traits.
This is easily solved by the crate [unimock], and using unimock = true
:
```rust use unimock::Unimock;
fn foo
fn bar
fn my_func(deps: &(impl Foo + Bar)) -> u32 { deps.foo() + deps.bar() }
fn main() { let deps = Unimock::new() .mock(|foo: &mut MockFoo| { foo.expectfoo().returning(|| 40); }) .mock(|bar: &mut MockBar| { bar.expectbar().returning(|| 2); });
assert_eq!(42, my_func(&deps));
} ```
unimock = true
implies mockall = true
.
Most often, you will only need to generate mock implementations in test code, and skip this for production code. For this, there are the test
variants:
mockall = test
unimock = test
which puts the corresponding attributes in #[cfg_attr(test, ...)]
:
```rust
fn foo
fn takes_foo(foo: impl Foo) {}
fn main() {
// we can still instantiate Unimock, but it's not useful,
// because now it doesn't implement Foo
:
let mock = unimock::Unimock::new();
//takes_foo(mock);
//--------- ^^^^ the trait Foo
is not implemented for Unimock
}
fn test() { // this compiles! let mock = unimock::Unimock::new(); takes_foo(mock); } ```
This is opt-in because there could be scenarios where this behaviour is not desired, e.g. when you write a library and want mocks exported for those.
License: MIT