--- "Randal L. Schwartz" <[EMAIL PROTECTED]> wrote: > >>>>> "James" == James Michael Dupont <[EMAIL PROTECTED]> writes: > > >> Good question. The scope of embedded rules in rmutt is dynamic > >> rather than lexical. In other words, embedded rules remain in > effect > >> > >> from the moment they are encountered in the expansion process > until > >> either the termination of the program or they are shadowed by > another > >> > >> embedded rule with the same name. > > Ahh, that makes the implementation so much easier, but also really > messes with things in a shared environment unless I make a copy of > tree before I start unwinding it. > > Otherwise, I was going to have to start adding "at end of rule, pop > the following entries", like Perl's "local" implementation. > > Now to try to figure out what the rest of his docs *really* mean. Of > course, I always have the source code. :) >
My interest is to create a file format that can be dumped by bison in the "compiler y file" stage, that like the vcg dump, gives you a full grammer, but in a form that can be fed to a tool like rmutt or Spew. This should be usable for generating test cases. Even if I need to feed it files as a test case, to present it random values of some rules. This would be great for creating test programs that cover all of the grammer of one programming language. On that can generate at least one test case per production.... Anyway, it might be needed to generate thousands of files, on per production, that will help test each one and contain only the rules needed for each one. mike ===== James Michael DuPont http://introspector.sourceforge.net/ __________________________________________________ Do you Yahoo!? HotJobs - Search new jobs daily now http://hotjobs.yahoo.com/
