Okay,
I just submitted one more fix. This time, I left in what we had done;
i.e. switched to flex -l, but I also added a test after this to check
for yylineno and use the old idea of just having it available in
Network.C if it doesn't exist. This takes care of the problem with
lexers we don't know, and also switches flex to add usable
information with yylineno. The speed is not so critical because this
only runs when networks are loaded.
David
On Thu, Jul 03, 2003 at 03:16:26PM -0600, David Thompson wrote:
So why don't I go back to the original plan and just see if I can
make configure detect if the lexer adds a yylineno? Wouldn't this be
a heck of a lot easier?
Maybe, but it's not what we want. The code (prior to your 'flex -l'
checkin) was syntactically correct--yylineno was properly defined. But
flex didn't make any use of it, ie. didn't update it as intended.
I don't know whether the code in question is very critical to
performance. If it isn't, the 'flex -l' fix is perfectly fine.
Regards,
Daniel.
--
.............................................................................
David L. Thompson Visualization and Imagery Solutions, Inc.
mailto:[EMAIL PROTECTED] 5515 Skyway Drive, Missoula, MT 59804
Phone : (406)756-7472
