As you said, doing this for all the classes (which you hate), is probably the most intutive way, especially if you are trying to do something throuh prototypeing.
The best would be if you could add code to the classes generated by hibernate and if you try to reverse engineer again it would leave the code you have added to the class alone.
The tools would then have to check if there already is a class with the same name in existence as it is about to generate, and if so only add methods/variables that doesn't already exist.
This is especially important if you are going to implement interfaces, then you probably would want access to the methods generated by Hibernate, which you won't be able to if Hibernate is just extending a base class.
I don't know how easy this is to implement but it would make the whole business completely intuitive and you wouldn't end up haveing 100+ or 200 classes for 100 tables, you would have 100 classes and thats it.
Any thoughts about this idea?
|