On 14 Dec 2006, at 21:49, Joel E. Denny wrote:
But by this idea, language options should then not be alterable by
startup
arguments. In particular, %language should only not be duplicated
by a
--language. Perhaps this idea is what you sense intuitively.
I'm afraid I don't know what is considered a "language option"
versus a
"compiler option".
Though an intuitive concept, subject to interpretation, a language
option would be the stuff in .y that should be present on any
compiler, though there is only one, Bison. :-) A compiler option is
what is compiler specific, like the #pragma in C/C++ preprocessing.
The language option is limited to things that are strictly needed to
generate the parser.
For example, --verbose seem to be a compiler option, the %verbose
that already exists should then not be there. Unless one can think of
a situation where the .output files is used in a parser or program.
On the other hand, %defines writes headers that are crucial for
getting the generated parser to work, so then --defines seem
unnecessary.
Or so, I am playing around.
Also, separate input files sounds like a paradigm
shift (and a lot of implementation work) that I don't think we're
ready
for.
Sure, and therefore probably not implementable. But the idea can be
used to think about it.
Hans Aberg
