下面是重写equals()操作时的一种处理方法。

public abstract class Section<T extends Section> extends Element<Section<T>> {
    Object attr1;

    /**
    * Compare one section object to another.
    *
    * @param obj the object being compared with this section object
    * @return true if this section and the other section are of the same
    * sub-class of section and their component fields are the same, false
    * otherwise
    */       
    @Override
    public boolean equals(Object obj) {
        if (obj == null) {
            // this exists, but obj doesn't, so they can't be equal!
            return false;
        }

        // prepare to cast...
        Section<?> other;

        if (getClass() != obj.getClass()) {
            // looks like we're comparing apples to oranges
            return false;
        } else {
            // it must be safe to make that cast!
            other = (Section<?>) obj;
        }

        // and then I compare attributes between this and other
        return this.attr1.equals(other.attr1);
    }
}

这似乎在Java 8中工作(甚至用-Xlint:unchecked编译)

计算机科学中的几乎每一个问题都可以通过添加一层间接方法来解决。

因此，引入一个非泛型对象，它的级别高于Map。如果没有上下文，它看起来不会很有说服力，但无论如何:

public final class Items implements java.io.Serializable {
    private static final long serialVersionUID = 1L;
    private Map<String,String> map;
    public Items(Map<String,String> map) {
        this.map = New.immutableMap(map);
    }
    public Map<String,String> getMap() {
        return map;
    }
    @Override public String toString() {
        return map.toString();
    }
}

public final class New {
    public static <K,V> Map<K,V> immutableMap(
        Map<? extends K, ? extends V> original
    ) {
        // ... optimise as you wish...
        return Collections.unmodifiableMap(
            new HashMap<String,String>(original)
        );
    }
}

static Map<String, String> getItems(HttpSession session) {
    Items items = (Items)
        session.getAttribute("attributeKey");
    return items.getMap();
}

*除了过多的间接层次。

在强制转换之前检查一下。

Object someObject = session.getAttribute("attributeKey");
if(someObject instanceof HashMap)
HashMap<String, String> theHash = (HashMap<String, String>)someObject;  

对于任何提问的人来说，收到你不确定类型的对象是很常见的。大量遗留的“SOA”实现传递各种对象，您不应该总是信任这些对象。(恐怖!)

EDIT修改了示例代码一次，以匹配海报的更新，在一些评论之后，我看到instanceof不能很好地使用泛型。然而，更改检查以验证外部对象似乎可以很好地使用命令行编译器。修订的例子现在发布。

下面是一个简短的示例，通过使用其他回答中提到的两种策略来避免“unchecked cast”警告。

Pass down the Class of the type of interest as a parameter at runtime (Class<T> inputElementClazz). Then you can use: inputElementClazz.cast(anyObject); For type casting of a Collection, use the wildcard ? instead of a generic type T to acknowledge that you indeed do not know what kind of objects to expect from the legacy code (Collection<?> unknownTypeCollection). After all, this is what the "unchecked cast" warning wants to tell us: We cannot be sure that we get a Collection<T>, so the honest thing to do is to use a Collection<?>. If absolutely needed, a collection of a known type can still be built (Collection<T> knownTypeCollection).

下面示例中的遗留代码接口在StructuredViewer中有一个属性“input”(StructuredViewer是一个树或表小部件，“input”是它背后的数据模型)。这个“输入”可以是任何类型的Java集合。

public void dragFinished(StructuredViewer structuredViewer, Class<T> inputElementClazz) {
    IStructuredSelection selection = (IStructuredSelection) structuredViewer.getSelection();
    // legacy code returns an Object from getFirstElement,
    // the developer knows/hopes it is of type inputElementClazz, but the compiler cannot know
    T firstElement = inputElementClazz.cast(selection.getFirstElement());

    // legacy code returns an object from getInput, so we deal with it as a Collection<?>
    Collection<?> unknownTypeCollection = (Collection<?>) structuredViewer.getInput();

    // for some operations we do not even need a collection with known types
    unknownTypeCollection.remove(firstElement);

    // nothing prevents us from building a Collection of a known type, should we really need one
    Collection<T> knownTypeCollection = new ArrayList<T>();
    for (Object object : unknownTypeCollection) {
        T aT = inputElementClazz.cast(object);
        knownTypeCollection.add(aT);
        System.out.println(aT.getClass());
    }

    structuredViewer.refresh();
}

当然，如果我们使用错误的数据类型的遗留代码(例如，如果我们将一个数组设置为StructuredViewer的“输入”而不是Java Collection)，上面的代码就会给出运行时错误。

调用方法的例子:

dragFinishedStrategy.dragFinished(viewer, Product.class);

在这种特殊情况下，我不会将Maps直接存储到HttpSession中，而是将我自己的类的实例存储到HttpSession中，而该类的实例又包含一个Map(类的实现细节)。然后可以确定映射中的元素是正确的类型。

但是如果你想检查Map的内容类型是否正确，你可以使用这样的代码:

public static void main(String[] args) {
    Map<String, Integer> map = new HashMap<String, Integer>();
    map.put("a", 1);
    map.put("b", 2);
    Object obj = map;

    Map<String, Integer> ok = safeCastMap(obj, String.class, Integer.class);
    Map<String, String> error = safeCastMap(obj, String.class, String.class);
}

@SuppressWarnings({"unchecked"})
public static <K, V> Map<K, V> safeCastMap(Object map, Class<K> keyType, Class<V> valueType) {
    checkMap(map);
    checkMapContents(keyType, valueType, (Map<?, ?>) map);
    return (Map<K, V>) map;
}

private static void checkMap(Object map) {
    checkType(Map.class, map);
}

private static <K, V> void checkMapContents(Class<K> keyType, Class<V> valueType, Map<?, ?> map) {
    for (Map.Entry<?, ?> entry : map.entrySet()) {
        checkType(keyType, entry.getKey());
        checkType(valueType, entry.getValue());
    }
}

private static <K> void checkType(Class<K> expectedType, Object obj) {
    if (!expectedType.isInstance(obj)) {
        throw new IllegalArgumentException("Expected " + expectedType + " but was " + obj.getClass() + ": " + obj);
    }
}

Two ways, one which avoids the tag completely, the other using a naughty but nice utility method. The problem is pre-genericised Collections... I believe the rule of thumb is: "cast objects one thing at a time" - what this means when trying to use raw classes in a genericised world is that because you don't know what is in this Map<?, ?> (and indeed the JVM might even find that it isn't even a Map!), it obvious when you think about it that you can't cast it. If you had a Map<String, ?> map2 then HashSet<String> keys = (HashSet<String>)map2.keySet() does not give you a warning, despite this being an "act of faith" for the compiler (because it might turn out to be a TreeSet)... but it is only a single act of faith. PS to the objection that iterating as in my first way "is boring" and "takes time", the answer is "no pain no gain": a genericised collection is guaranteed to contain Map.Entry<String, String>s, and nothing else. You have to pay for this guarantee. When using generics systematically this payment, beautifully, takes the form of coding compliance, not machine time! One school of thought might say that you should set Eclipse's settings to make such unchecked casts errors, rather than warnings. In that case you would have to use my first way.

package scratchpad;

import java.util.HashMap;
import java.util.Iterator;
import java.util.Map;
import java.util.Vector;

public class YellowMouse {

    // First way

    Map<String, String> getHashMapStudiouslyAvoidingSuppressTag(HttpSession session) {
      Map<?, ?> theHash = (Map<?, ?>)session.getAttribute("attributeKey");

      Map<String, String> yellowMouse = new HashMap<String, String>();
      for( Map.Entry<?, ?> entry : theHash.entrySet() ){
        yellowMouse.put( (String)entry.getKey(), (String)entry.getValue() );
      }

      return yellowMouse;
    }


    // Second way

    Map<String, String> getHashMapUsingNaughtyButNiceUtilityMethod(HttpSession session) {
      return uncheckedCast( session.getAttribute("attributeKey") );
    }


    // NB this is a utility method which should be kept in your utility library. If you do that it will
    // be the *only* time in your entire life that you will have to use this particular tag!!

    @SuppressWarnings({ "unchecked" })
    public static synchronized <T> T uncheckedCast(Object obj) {
        return (T) obj;
    }


}