将最后一项推入“可观察的"(序列) [英] Take the last item pushed to an Observable (Sequence)
问题描述
我在一个类中有一个IObservable<Item>
,我想公开一个只读属性,该属性提供在给定时间推送到可观察对象的最后一项.因此它将提供单个值Item
.
I have an IObservable<Item>
inside a class and I want to expose a read-only property that provides the last item pushed to the observable at a given time. So it will provide a single value of Item
.
如果未推送任何值,则必须返回默认值.
If no value has been pushed, then it will have to return a default value.
我如何做到这一点而不必订阅可观察的内容并拥有后备字段"?
How can I do this without having to subscribe to the observable and having a "backing field"?
推荐答案
仅在这里补充@Asti的答案,也许可以帮助您避免沮丧:
Just to supplement @Asti's answer a bit here, and perhaps help you with your frustration:
可观察的事物不是物理的事物",而是更具逻辑性的概念.通常将Rx与LINQ进行比较,并且在很多时候,这是一个公平的比较.但是,当您开始谈论数据结构时,它就会崩溃:出于学习目的,LINQ的可枚举与Lists足够相似.
An observable isn't a physical 'thing', it's more a logical concept. Rx is often compared to LINQ, and it's a fair comparison much of the time. It breaks down though when you start talking data structures: LINQ's enumerables are similar enough to Lists for learning purposes.
但是,在Rx端,根本没有与List相当的产品.一个可观察的是一个瞬态数据结构,所有运算符都处理这种瞬态状态.如果您要寻找永久状态,就离开Rx.
However, on the Rx side, there's simply no good equivalent to List. An observable is a transient data structure, all operators deal with this transient state. If you're looking for a permanent state, you're leaving Rx.
已经说过,将可观察到的状态转换为某种状态是一个常见的问题,并且有一些程序包可能会对您有所帮助:ReactiveUI可能是最著名的. ReactiveProperty是另一个.这两个软件包都有缺陷,但可能会对您有所帮助.
Having said that, converting an observable to some sort of state is a common problem, and there are some packages that may help you: ReactiveUI is perhaps the most known. ReactiveProperty is another. Both of these packages are flawed, but may help you.
如果您只是在寻找一种简单的方法来获得背景字段,而无需锅炉电镀背景字段,那么这将起作用:
If you're simply looking for an easier way to get a backing field, without boiler-plating out a backing field, this will work:
public static class ReactivePropertyExtensions
{
public static ReactiveProperty<T> ToReactiveProperty<T>(this IObservable<T> source)
{
return new ReactiveProperty<T>(source);
}
public static ReactiveProperty<T> ToReactiveProperty<T>(this IObservable<T> source, T defaultValue)
{
return new ReactiveProperty<T>(source, defaultValue);
}
}
public class ReactiveProperty<T> : IDisposable
{
private IObservable<T> Source { get; }
private IDisposable Subscription { get; }
public T Value { get; private set; }
public ReactiveProperty(IObservable<T> source)
: this(source, default(T)) { }
public ReactiveProperty(IObservable<T> source, T defaultValue)
{
Value = defaultValue;
Source = source;
Subscription = source.Subscribe(t => Value = t);
}
public void Dispose()
{
Subscription.Dispose();
}
}
示例用法:
var ticker = Observable.Interval(TimeSpan.FromSeconds(1))
.Publish().RefCount();
var latestTickerValue = ticker.ToReactiveProperty();
Console.WriteLine(latestTickerValue.Value);
await Task.Delay(TimeSpan.FromSeconds(1));
Console.WriteLine(latestTickerValue.Value);
await Task.Delay(TimeSpan.FromSeconds(3));
Console.WriteLine(latestTickerValue.Value);
这篇关于将最后一项推入“可观察的"(序列)的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持IT屋!