Re: inserting into tab.h, after %union

[Jeff Inman]

On Dec 6, 2006, at 2:58 AM, Joel E. Denny wrote:
> On Wed, 6 Dec 2006, Hans Aberg wrote:
>
> > > Potentially, but I'm not inclined to complicate the existing  
> > > directives as
> > > we grope in the dark trying to accommodate future target  
> > > languages that we
> > > haven't even identified yet.
> >
> > This is what I am saying: you may already have done that by the  
> > current
> > commands.
>
> As far as these directives being complicated, I just don't think  
> they are.
> Neither did Akim Demaille (who initially proposed the names % 
> provides and
> %requires), Paul Eggert, Paolo Bonzini, and Pupeno when they first  
> heard
> of them.  Two months later, Paul said he had forgotten what the  
> directives
> meant, but he made no further comment.  Jeff Inman's discomfort seems
> quite mild, but I'd be glad to hear further discussion from him.

Just to chime in, then ... as with any software tool, if there are  
just a few awkward
terms to learn, I can learn what I need to know from a manual, and re- 
read it
when I have to do maintenance, until the terms just become  
"natural".  The
syntax you've described does seem kinda awkward to me, but at least  
it solves
my problem.  I would think that %declares{...} and %defines{...}  
might be a little
more clear in multiple languages, but you guys have been looking at  
this more
than I have.

I like that you are proposing specific syntax for specific functions,  
instead of having
context-sensitive interpretations of simpler syntax.  The proposed  
way seems to me,
at first glance, to make things simpler to understand.  The language  
*does* seem to be
increasing in complexity, with special-case interpretations for C++  
versus C, etc.
From a "marketing" point of view, that seems to me like a dangerous  
trend for bison.
The more hairy it becomes, the more pressure there is for someone to  
come up with
a "simpler" tool to replace it.  (good luck, I know.)  Next thing you  
know, they make
something prettier, but they also make whitespace significant, and no  
good language
does that!


OFF TOPIC: is there a simple way to use a C++ skeleton but to have  
the token
enumerants generated directly into the std namespace?   
yy::parser::token::FOO
is a mouthful.  Maybe I could get halfway there with %name-prefix="std".



Regards,

Jeff




Jeff Inman
Parallel and Distributed Visualization
Los Alamos National Laboratory
PO Box 1663, B287 / Los Alamos, NM 87545