On Tue, 20 Sep 2005, Paul Eggert wrote: > "Joel E. Denny" <[EMAIL PROTECTED]> writes: > > > You seem to have missed the yylval and yylloc occurrences I reported in > > local.at and calc.at. > > I tried changing those, and the tests stopped working. I suppose we
What do you mean by `stopped working'? I changed them and all the tests continued to pass. Are these false results? Note that the changes in calc.at depend on the changes in local.at. > can call these usages of yylval/yylloc as being tests of internal > facilities (though I admit that I didn't think about it much). I disagree. There's no reason for parameters to yyerror() and to yylex() to intrude on the parser's namespace. I've noticed that the bison documentation is careful about this. Joel
