在Java中，在处理序列化和来自另一个可实例化类的引用时，引用用户定义的类如何工作？

Question

I have a class which is serializable, say class " WriteMe ". This class needs another class to update one of its members. Said class is relatively large, since WriteMe is just a wrapper class mostly. We will call the large class " MegaClass ". Therefore we have something like this:

public class WriteMe implements Serializable
{
    private ArrayList<String> list;
    private String name;

    public WriteMe(String s)
    {
        name = s;
    }

    public void update(MegaClass m)
    {
        list = m.getList(name);
    }
}

The hopeful point in doing this is that I don't want to store MegaClass as a class variable like the ArrayList and String are. I'm hoping by passing it in to a method and only using it there where it's needed , I will reduce the amount of memory used when creating a WriteMe , and also reduce the file-size of a WriteMe object, if I were to pass it to an ObjectOutputStream or similar (so that it can be written to file).

Does " Megaclass m " just store a pointer to the instantiation of that object? If so, I would imagine that it doesn't matter if I make it a class variable or not, since an address is just a few bytes or so. Is there any advantage to using the object passed into the method only, as shown above? The other way is to pass it into the constructor and assign as a class variable. Is either way advantageous?

Answer 1

As Mark W. suggests, the best thing to do is to separate the use of "MegaClass" and "WriteMe", so that all that is required by WriteMe is the data.

public void update(ArrayList<String> newList)
{
    list = newList;
}

In this manner, a super class which has a reference to both classes may inquire from MegaClass "What is the correct data?", and pass this information to WriteMe! A bonus of this is that no transient data is needed, nor ANY reference to the MegaClass at all!