Re: symbol declarations after rules

[Paul Eggert]

"Joel E. Denny" <address@hidden> writes:

> To keep things less confusing, the user should be advised to do things 
> either the Yacc way or the new way with order-independent declarations.  

Yes, that makes sense.  How about requiring that?  I.e., report an
error if the user confusingly tries to use both methods.  That way, we
don't have to define what happens after you shoot yourself in the
foot.

>   before-header
>   ----------------------------
>   before-definitions
>   Bison-generated union, YYLTYPE, enum, classes
>   after-definitions
>   ----------------------------
>   after-header

This makes things much clearer, thanks.

One notational thought.  Perhaps we should change "before-definitions"
to "start-header", and "after-definitions" to "end-header"?  That
might help people remember better the relationships among the
%*-header directives.  Also, it will give them the pattern "%*-header"
to remember the concepts by.