Re: bison'd files don't compile under Microsoft VC++

[Hans Aberg]

At 15:42 -0800 2002/02/13, Paul Eggert wrote:
>> It is not so difficult to fix:
>
>That may well be, but given the latest hassle with Microsoft compilers
>I now tend to agree with Akim that it is not worth the aggravation.
>We've spent too much time worrying about C++ namespace cleanliness in
>what is basically a C parser, and it's causing more problems than it's
>curing.  We should defer worring about C++ namespace issues until we
>have a proper C++ parser.

You should make up your mind about the future of the compile C as C++
feature: If you think it will never be used once a proper C++ parser can be
generated, then there is no point in supporting it.

On the one hand, the dynamic stack that Akim wrote for use with C cannot be
used with C++ types that have non-trivial constructors. But on the other
hand, some people may want it.

I wrote my own proper C++ parser generating features, I once I started
using that, I decided to never use the compile C as C++ feature.

Nevertheless, if the compile C as C++ feature and proper C++ namespaces
should remain (even though if looks as though you do not want them
anymore), one might have two macros
  YYNO_NAMESPACE  // If defined, no C++ namespace std.
  YYC_HEADERS     // If defined, use C .h compatibility headers.

  Hans Aberg