pub trait Generator<ResumeArg = ()> {
type Yield;
type Return;
fn resume(
self: Pin<&mut Self>,
resume_arg: ResumeArg
) -> GeneratorState<Self::Yield, Self::Return>;
}
Expand description
The trait implemented by GeneratorFn
s.
Generators, also commonly referred to as coroutines, provide an ergonomic definition for iterators and other primitives, allowing to write iterators and iterator adapters in a much more imperative way, which may sometimes improve the readability of such iterators / iterator adapters.
Example
use ::next_gen::prelude::*;
fn main ()
{
#[generator(yield(i32))]
fn generator_fn ()
-> &'static str
{
yield_!(1);
return "foo";
}
mk_gen!(let mut generator = generator_fn());
let mut next = || generator.as_mut().resume(());
match next() {
| GeneratorState::Yielded(yielded) => assert_eq!(yielded, 1),
| GeneratorState::Returned(_) => panic!("unexpected return from resume"),
}
match next() {
| GeneratorState::Yielded(_) => panic!("unexpected yield from resume"),
| GeneratorState::Returned(returned) => assert_eq!(returned, "foo"),
}
}
Generator
vs. Iterator
-
a
Generator
can return a non-trivial value when exhausted, contrary to anIterator
, -
but they require to be
Pin
-ned in order to bepoll
ed.
We thus have the following impls:
impl<Item, R, F>
Iterator
for
Pin<&'_ mut (
GeneratorFn<Item, F, ()>
)>
where
GeneratorFn<Item, F, ()> : Generator<(), Yield = Item, Return = R>,
impl<Item, R>
Iterator
for
Pin<&'_ mut (
dyn '_ + Generator<(), Yield = Item, Return = R>
)>
and, when under #[cfg(feature = "alloc")]
, we also have:
impl<Item, R, F>
Iterator
for
Pin<Box<
GeneratorFn<Item, F, ()>,
>>
where
GeneratorFn<Item, F, ()> : Generator<(), Yield = Item, Return = R>,
impl<Item, R>
Iterator
for
Pin<Box<
dyn '_ + Generator<(), Yield = Item, Return = R>,
>>
A remark regarding the lack of blanket impl and coherence
Since
{,Into}Iterator
is defined in::core
, and this definition ofGenerator
is a third-party library one, for coherence reasons, it is not possible to implement{,Into}Iterator
for all theimpl Generator
s.That being said, the above impls do cover the
dyn Generator
andGeneratorFn
cases, which ought to cover all the#[generator] fn
-originating generator instances.Should such impls not be enough, there is always the
.boxed_gen_into_iter()
and.gen_into_iter()
methods to convert any pinned generator (providedResumeArg = ()
) into an interator.
Associated Types
The type of value this generator yields.
This associated type corresponds to the yield_!
expression and the
values which are allowed to be returned each time a generator yields.
For example an iterator-as-a-generator would likely have this type as
T
, the type being iterated over.
Required methods
Resumes the execution of this generator.
This function will resume execution of the generator or start execution
if it hasn’t already. This call will return back into the generator’s
last suspension point, resuming execution from the latest yield_!
.
The generator will continue executing until it either yields or returns,
at which point this function will return.
Return value
The GeneratorState
enum returned from this function indicates what
state the generator is in upon returning.
If the Yield
variant is returned then the
generator has reached a suspension point and a value has been yielded
out. Generators in this state are available for resumption at a later
point.
If Return
is returned then the generator has completely finished
with the value provided. It is invalid for the generator to be resumed
again.
Panics
This function may panic if it is called after the Return
variant has
been returned previously. While generator literals in the language are
guaranteed to panic on resuming after Return
, this is not guaranteed
for all implementations of the Generator
trait.