'getCurrentInputLocation': since it is expected to create an object, I don't like 'get'

My concern is that the function takes a stateful object as parameter, so I image it will return a view of the current state of the parser.
the word 'to' seems to me like a 'conversion'.
Not so important to me, I am very new to this code.

What do you mean by "state the expected behavior"?

For instance, if the method is called twice is it expected to return two equivalent objects, the same object, or what ?

Object a =  toInputLocation( parser );
Object b =  toInputLocation( parser );

should b == a, b.quals(a) or it is not important ?

Since the implementation is really something that will be specific to the user, and that the user will have to cast, I don't see the benefit of creating an empty interface

Fine