2022 -使用useEffect钩子

此时最好的选择是使用useEffect钩子。useEffect允许您设置一个函数，该函数可以修改状态以响应某些异步事件。debounging是异步的，因此useEffect可以很好地用于此目的。

如果从钩子返回一个函数，返回的函数将在钩子再次被调用之前被调用。这让您可以取消之前的超时，有效地解除函数。

例子

这里我们有两个状态，value和tempValue。设置tempValue将触发useEffect钩子，该钩子将启动一个1000ms超时，该超时将调用一个函数将tempValue复制为value。

钩子返回一个取消定时器设置的函数。当钩子再次被调用时(即按下另一个键)，超时被取消并重置。

const DebounceDemo = () => {
  const [value, setValue] = useState();
  const [tempValue, setTempValue] = useState();

  // This hook will set a 1000 ms timer to copy tempValue into value
  // If the hook is called again, the timer will be cancelled
  // This creates a debounce
  useEffect(
    () => {
      // Wait 1000ms before copying the value of tempValue into value;
      const timeout = setTimeout(() => {
        setValue(tempValue);
      }, 1000);

      // If the hook is called again, cancel the previous timeout
      // This creates a debounce instead of a delay
      return () => clearTimeout(timeout);
    },
    // Run the hook every time the user makes a keystroke
    [tempValue]
  )

  // Here we create an input to set tempValue. 
  // value will be updated 1000 ms after the hook is last called, 
  // i.e after the last user keystroke.
  return (
    <>
      <input 
        onChange={
          ({ target }) => setTempValue(target.value)
        }
      />
      <p>{ value }</p>
    </>
  )
}

Julen的解决方案有点难以阅读，这里有更清晰和准确的反应代码，供那些根据标题而不是问题的微小细节绊倒他的人使用。

Tl;dr版本:当你要更新到观察者发送调用一个调度方法，反过来实际上会通知观察者(或执行ajax等)

使用示例组件jsfiddle完成jsfiddle

var InputField = React.createClass({

    getDefaultProps: function () {
        return {
            initialValue: '',
            onChange: null
        };
    },

    getInitialState: function () {
        return {
            value: this.props.initialValue
        };
    },

    render: function () {
        var state = this.state;
        return (
            <input type="text"
                   value={state.value}
                   onChange={this.onVolatileChange} />
        );
    },

    onVolatileChange: function (event) {
        this.setState({ 
            value: event.target.value 
        });

        this.scheduleChange();
    },

    scheduleChange: _.debounce(function () {
        this.onChange();
    }, 250),

    onChange: function () {
        var props = this.props;
        if (props.onChange != null) {
            props.onChange.call(this, this.state.value)
        }
    },

});

这个解决方案不需要任何额外的库，它也会在用户按下enter时启动:

const debounce = (fn, delay) => {
    let timer = null;
    return function() {
        const context = this,
        args = arguments;
        clearTimeout(timer);
        timer = setTimeout(() => {
            fn.apply(context, args);
        }, delay);
    };  
}

const [search, setSearch] = useState('');
const [searchFor, setSearchFor] = useState(search);

useEffect(() => {
    console.log("Search:", searchFor);
}, [searchFor]);

const fireChange = event => {
    const { keyCode } = event;
    if (keyCode === 13) {
        event.preventDefault();
        setSearchFor(search);
    }
}

const changeSearch = event => {
    const { value } = event.target;
    setSearch(value);
    debounceSetSearchFor(value);
};

const debounceSetSearchFor = useCallback(debounce(function(value) {
    setSearchFor(value);
}, 250), []);

输入可以是这样的:

<input value={search} onKeyDown={fireChange} onChange={changeSearch}  />

与其在debounce()中包装handleOnChange，不如在debounce()中包装回调函数中的ajax调用，从而不破坏事件对象。就像这样:

handleOnChange: function (event) {
   debounce(
     $.ajax({})
  , 250);
}

I was searching for a solution to the same problem and came across this thread as well as some others but they had the same problem: if you are trying to do a handleOnChange function and you need the value from an event target, you will get cannot read property value of null or some such error. In my case, I also needed to preserve the context of this inside the debounced function since I'm executing a fluxible action. Here's my solution, it works well for my use case so I'm leaving it here in case anyone comes across this thread:

// at top of file:
var myAction = require('../actions/someAction');

// inside React.createClass({...});

handleOnChange: function (event) {
    var value = event.target.value;
    var doAction = _.curry(this.context.executeAction, 2);

    // only one parameter gets passed into the curried function,
    // so the function passed as the first parameter to _.curry()
    // will not be executed until the second parameter is passed
    // which happens in the next function that is wrapped in _.debounce()
    debouncedOnChange(doAction(myAction), value);
},

debouncedOnChange: _.debounce(function(action, value) {
    action(value);
}, 300)

我在这个问题下找不到任何答案，提到我正在使用的方法，所以只想在这里提供一个替代解决方案，我认为这是最适合我的用例。

如果您正在使用流行的react钩子工具包库react-use，那么有一个名为useDebounce()的实用工具钩子，它以一种相当优雅的方式实现了谴责逻辑。

const [query, setQuery] = useState('');

useDebounce(
  () => {
    emitYourOnDebouncedSearchEvent(query);
  },
  2000,
  [query]
);

return <input onChange={({ currentTarget }) => setQuery(currentTarget.value)} />

有关详细信息，请直接检查库的github页面。

https://github.com/streamich/react-use/blob/master/docs/useDebounce.md