我在这种情况下遇到了一个问题(这里的代码是可构建的):
extern crate rand;
use rand::{Isaac64Rng, SeedableRng, Rng};
pub trait GeneticAlgorithm<R, Ins, C> : Clone where R: Rng {
fn mate(parents: (&Self, &Self), rng: &mut R) -> Self;
fn mutate<F>(&mut self, rng: &mut R, mutator: F) where F: FnMut(&mut Ins);
fn call<F>(&self, program: F) where F: FnOnce(&C);
}
pub struct Mep<Ins> {
instructions: Vec<Ins>,
unit_mutate_size: usize,
crossover_points: usize,
}
impl<Ins> Mep<Ins> {
//Generates a new Mep with a particular size and takes a closure to generate random instructions
pub fn new<I>(unit_mutate_size: usize, crossover_points: usize, instruction_iter: I) -> Mep<Ins>
where I: Iterator<Item=Ins> {
Mep{instructions: instruction_iter.collect(), unit_mutate_size: unit_mutate_size,
crossover_points: crossover_points}
}
}
impl<Ins> Clone for Mep<Ins>
where Ins: Clone {
fn clone(&self) -> Self {
Mep{instructions: self.instructions.clone(), unit_mutate_size: self.unit_mutate_size,
crossover_points: self.crossover_points}
}
}
impl<R, Ins> GeneticAlgorithm<R, Ins, Vec<Ins>> for Mep<Ins> where R: Rng, Ins: Clone {
fn mate(parents: (&Mep<Ins>, &Mep<Ins>), rng: &mut R) -> Mep<Ins> {}
fn mutate<F>(&mut self, rng: &mut R, mut mutator: F) where F: FnMut(&mut Ins) {}
fn call<F>(&self, program: F) where F: FnOnce(&Vec<Ins>) {
program(&self.instructions);
}
}
fn main() {
let mut rng = Isaac64Rng::from_seed(&[1, 2, 3, 4]);
let (a, b) = {
let mut clos = || Mep::new(3, 3, rng.gen_iter::<u32>().map(|x| x % 10).take(10));
(clos(), clos())
};
let mut c = Mep::mate((&a, &b), &mut rng);
c.mutate(&mut rng, |ins: &mut u32| *ins = 2);
c.call(|x: &Vec<u32>| panic!());
}
Rust声称它无法在某处推断出某种类型,但我不知道如果这是问题,如何指定闭包的类型,我也无法确定哪个特定的泛型参数导致了这个问题:
main.rs:48:7: 48:36 error: unable to infer enough type information about `_`; type annotations or generic parameter binding required [E0282]
main.rs:48 c.call(|x: &Vec<u32>| panic!());
^~~~~~~~~~~~~~~~~~~~~~~~~~~~~
需要指定哪个通用参数以及如何确定?如果无法推断,如何指定预期特征:GeneticAlgorithm<Isaac64Rng, u32, Vec<u32>>
如果有人想自己构建原始代码,我就是EntityFieldQuery Extra Fields。
答案 0 :(得分:1)
impl<R, Ins> GeneticAlgorithm<R, Ins, Vec<Ins>> for Mep<Ins> where R: Rng, Ins: Clone {
// ...
}
对于impl
的所有可能值,此GeneticAlgorithm
块为Mep<Ins>
实施R
。这意味着特定GeneticAlgorithm
的{{1}}特征有多个实现。当您调用Mep<Ins>
和mate
方法时,编译器能够从参数中解析特定的实现,但是当您调用mutate
时,编译器无法解析特定的实现,因为call
不受限制。
要解决此问题,请将R
通用参数移至R
和mate
方法。
mutate