[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Two questions for scanning/parsing, IDENT vs. TYPE, operators


I would like to be able to let the scanner descide if a recognized token 
is simply a IDENT or a TYPE.

The scanner could simply lookup this in a symbol table.

Thus I would like to have a SymbolTable, which is shared between
scanner and parser.

Now I would need the ability to perform actions during parsing, e.g.
placing an ident into the symboltable thus letting the scanner know,
that I would like to get a TYPE token the next time this ident is

How do I define my own token classes and let the parser and scanner 
interact in this way?

Very similar: I would like to have user defined operators.
This are not only the standard operators like in C++ or in JAVA.
E.G. the user should be able to do this (in C++ style): 

	Result operator ** (Number lhs, Number rhs); // XX

This means I want to be able to tell the scanner that in some
circumstances ** is a operator. E.G: a USEROP or so, as I do not know
what kind of operators the USER will define in its program I want
to have an action which tells the scanner that now an operator
like that above in // XX is available. Later I want to remove this
operator eventually.

Thinking I have several token categories:

I would like to be able to extend each of them during parsing.

I would like to get back from the scanner IDENT(String),
BRACKET('(') or BRACKET ('{') or OPERATOR('**') etc.

Thanx for your help.


Angelo Schneider           OOAD/UML           Angelo.Schneider@xcc.de
Putlitzstr. 24         Patterns/FrameWorks       Fon: +49 721 9812465
76137 Karlsruhe             C++/JAVA             Fax: +49 721 9812467