Re: shift/reduce conflict with unary

[Derek M Jones]

cwcaceres,

> > I take it the shift reduce conflict occurs on the input:
> >
> > NUMBER PLUS NUMBER potentially-other-tokens
> >
> > which can be reduced to a single summation expression
> > or shifted because there may be a following UPLUS
> >
> > Your problem is that you are trying to do too much in too few rules.
> >
> > Create some productions for the unary operators and separate ones for
> > the binary operators.  See how the C standard has done it, for an
> > example.
> Sorry, I don't understand. How would I create separate productions for unary
> and binary operators?
>
> I looked at the C++ grammar. It seems somewhat upside down, but rearranging
> I think it's a bit similar to what I did which is why I'm still confused.

Don't forget that your parser has to operate with a single token of
look-ahead.  You think it is similar because you are taking a global
view (the two grammars do specify very similar token sequences).

Try to think in terms of a push down automata that has to decide what
to do given its current state and the next token.  In the case of
the C++ grammar there are no shift/reduce decision conflicts (I'm
assuming you copied it correctly).

--
Derek M. Jones                              tel: +44 (0) 1252 520 667
Knowledge Software Ltd                      mailto:address@hidden
Applications Standards Conformance Testing    http://www.knosof.co.uk