for what purpose do abstracting nodes mark attributes in abstract syntax trees?
This is done because of the fact that they create generic objects of nodes, and not objects of separate strictly-typed classes? Or the use of attributes makes it somehow easier to write generalized algorithms with tree nodes based on the values of the attributes? Or are the member variables of the node-classes of the parse tree just these very attributes, is this just another terminology?
In general, what should be read (electronic, available on the Internet, in vain) for clarity to come?
