Java编译器为什么不将其标记为错误？

Question

I have created an enumerated type in my Java webapp as a way of abstracting the procedure of adding objects to the session attributes. The idea being to precipitate an error if I try and map the wrong object type to a particular key. The enum code looks like this:

public enum SessionVariables
{
CURRENT_BRIEF {

    @Override
    public <Briefable> void setValue(Briefable item)
    {
        FacesContext.getCurrentInstance().getExternalContext().getSessionMap().put(this.toString(), item);
    }

    @Override
    @SuppressWarnings("unchecked")
    public <Briefable> Briefable getValue()
    {
        return (Briefable) FacesContext.getCurrentInstance().getExternalContext().getSessionMap().get(this.toString());
    }
};

public abstract <T> void setValue(T item);
public abstract <T> T getValue();
}

All fine so far, no compilation problems.

However, if I actually try and use the methods with a wrong type, no error is found by the compiler, for instance:

SessionVariables.CURRENT_BRIEF.setValue("String value");

compiles normally.

Any ideas why this is the case? Have I missed something out? I'm using Java SE 8, EE7 on a Windows 7 machine.

Answer 1

The declared type of CURRENT_BRIEF is SessionVariables . That type has the method <T> void setValue(T item) and your method invocation complies with that signature.

Furthermore, when you wrote

public <Briefable> void setValue(Briefable item)

you basically restated the same thing, just naming your type parameter Briefable . The actual type will still be inferred from the calling context, which means that Briefable resolves to String for that invocation.