Re: [Plantool-developer] Flex errors and warnings

[Maurits v. R. <ma...@va...>]

Lomwel, Alexander-van wrote:
> Reinout;
> has a good approach here. Lets stick to it. 
> 
> The root couse analysis is flawed however. flex creates from the tokeniser
> files a flex.c file. Herein some defines steer the compilation like
> YY_NO_UNPUT, and some useless cose is automaticly inserted like the command
> yy_flex_realloc. (read BUG) These result in compiler warnings, which I
> sought to surpress, but which cause problems with older versions of flex,
> which don't seem to have the BUG, but also not know what to do with the
> define.

I searched a bit on internet and it seems to be a known bug, bugging 
more people. :-(

> PS Overal conclusion:
> we live with the warnings.
> Get rid of #define YY_NO_UNPUT, and the function CatchSillyWarnings().
> 	As MvR already did for the most part.

I committed this.


-- 
Maurits van Rees | ma...@va...
http://maurits.vanrees.org/
"Peace is not the absence of war; it is the absence of fear."