Difference between getClass().getClassLoader().getResource() and getClass.getResource()?

The second one calls the first one. The difference is described in the javadoc.

The first one takes paths that don't start with a /, and always starts at the root of the classpath.

The second one takes path that can start with a /. If it does, it starts at the root of the classpath. If not, it starts at the package of the class on which the method is called.

So getClass().getClassLoader().getResource("foo/bar.txt") is equivalent to getClass().getResource("/foo/bar.txt").

And, assuming getClass() returns a class that is in the package foo, getClass().getResource("bar.txt") is equivalent to getClass().getClassLoader().getResource("foo/bar.txt")

which one should I use in what circumstances?

Neither. You should call Thread.currentThread().getContextClassLoader().

This has the benefit of not needing to be changed depending on whether you're calling from a static or instance method.

And more important, it will properly handle classloader delegation inside a container. If you don't use it, you might find that an "application" resource isn't found because the class that's loading it was loaded by a classloader higher up the delegation hierarchy.

Basically, Class.getResource() allows you to specify a path relative to the package of the class, whereas ClassLoader.getResource() is always an "absolute" path.

So:

foo.bar.Baz.class.getResource("data.txt")

is equivalent to:

some.Other.class.getResource("/foo/bar/data.txt")

and they're both equivalent to:

some.Other.class.getClassLoader().getResource("foo/bar/data.txt")

(Assuming some.Other and foo.bar.Baz are loaded by the same classloader, of course.)