[EMAIL PROTECTED] writes: > the added YYFPUTC, YYFPUTS in glr.c > makes redirection of the debug output > to another stream more complicated.
In what way? All you need to do is define YYFPRINTF, i.e., do the same thing you were already doing before the new macros were added. > would prefer easy redirection of the > debug stream above fastness of fputc > compared to printf. If we also wanted higher performance in the case of redirected debug streams, we'd need to let the user be able to define YYFPUTC and YYFPUTS. Yes, that would complicate things along the lines that you were probably thinking. But is there a real need for this?
