On Tue, 2007-05-01 at 19:34 +0200, Francois Pottier wrote:
> > So basically, menhir is saying "what happens if you hit
> > end of stream here? What should be done?" and the answer
> > is "it can't happen".
> 
> What I am trying to say is, Menhir doesn't ask this question. It asks:
> "what happens if I reach that state? should I request a lookahead token
> from the lexer, or shouldn't I?"

Perhaps I do not fully understand but:

I do not use 'eof' (end-of-stream pseudo token).
I get this for an unused token:

"Warning: the token WHENCE is unused."

and I should get the same for eof IMHO (but possibly 
suppressed message since it's not a user defined
token).

What i actually get is 145 end-of-stream conflicts .. 
but i get no conflicts on WHENCE.

That is, instead of treating the token set precisely as the
set of user tokens + eof, menhir is treating eof specially
and not consistently with other tokens.

There's no conflict: in every one of these states if 'eof'
turns up its a syntax error, exactly the same as if WHENCE
turns up. 

I'm not sure i fully understand the 'do we need lookahead'
issue: I would have thought: you need a fetch if your action is 
shift, and not if it is a reduce.



-- 
John Skaller <skaller at users dot sf dot net>
Felix, successor to C++: http://felix.sf.net