Re: assert failure at line 1896

[Derek M Jones]

Joel,

> Replacing this in your grammar:
>
>   #define YYSTYPE char *
>
> with this:
>
>   #define YYSTYPE YYSTYPE
>   typedef char *YYSTYPE;
>
> should work around the bug for now.
>
> Akim, I believe the const in b4_yysymprint_generate was your commit.  Do 
> you want to require users to typedef YYSTYPE, or should we remove the 
> const?

Why not just switch the order of "const YYSTYPE" to "YYSTYPE const".
It looks a bit odd, but the semantics work for both typedef and
macro definition.

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