CODEWITHSUNDEEP
javaclass
Chop
Chop

Reputation: 4517

Class.forName() versus direct class loading

I found some code loading resources this way:

Class.forName(myClass.class.getName()).getResourceAsStream("myResource");

First, I wondered about the interest of using such a structure. It appears Class.forName("className") enables a dynamic loading, loading the class only when needed (this is the typical structure when loading a JDBC driver, for instance).

Yet, is not the dynamic loading inefficient in this case since the class is in code? Would there be any difference if I wrote the following?

myClass.class.getResourceAsStream("myResource");

Upvotes: 3

Views: 350

Answers (2)

Stephen C
Stephen C

Reputation: 718678

Your question doesn't make a lot of sense as written. Class.forName and Class.getResourceAsStream are doing different things. They are interchangeable.

The only way to compare them would be if you were comparing loading a class full of (say) static fields initialized with data versus reading a resource containing the same data. If that is what you are talking better, the getResourceAsStream approach is better in a number of respects:

  • there are no limits on the amount of data you can read that way,
  • you don't have to embed the data in "code",
  • your memory footprint will be smaller,
  • you can load just the subset of the data that you require,
  • it is probably faster,
  • and so on.

If you are asking if you can use getResourceAsStream() to load classes (in the normal sense), the answer is no. Sure you can get the bytecodes, but you have to do "other stuff" to turn those bytecodes into executable classes / methods in the JVM.

Upvotes: 0

popfalushi
popfalushi

Reputation: 1362

mmyClass.class.getResourceAsStream("myResource"); should be prefered option since it does not make search. Result is the same since myClass.class.getName() is used in forName, not just predefined string.

Upvotes: 1

Related Questions