Re: %destructor feedback

[Joel E. Denny]

On Wed, 7 Dec 2005, Akim Demaille wrote:

> > > > "Joel" == Joel E Denny <address@hidden> writes:
>
> > On Wed, 7 Dec 2005, Akim Demaille wrote:
> >>>>> "Paul" == Paul Eggert <address@hidden> writes:
> >>
> >> > Why can't we simply rewrite the code so that it doesn't declare
> >> > unused variables?
> >>
> >> Because you don't know whether yyerror, for instance, will use the
> >> location, or the %parse-param.  But some users need them there.
>
> > But the user writes yyerror, and I don't believe YYUSE is documented for
> > the user to invoke.
>
> Right :)  Consider the %printers and %destructors for instance.

In those cases, it also seems reasonable to leave it in the hands of the 
user.  That is, if the user doesn't reference a semantic value in its 
printer or destructor, why did he declare its printer or destructor?  Is 
this a legitimate usage?  Shouldn't he be warned *somehow*?  Maybe even by 
bison?  Whether $$ occurs can be computed.

In general it seems that, as long as the interface for a passage of user 
code is documented, any compiler or lint warnings for that code should be 
the user's problem.

Note that I'm just talking philosophy here.  I realize the implementation 
of this might require an internal redesign, as Paul said.

Joel