On Wed, Oct 13, 2010 at 2:00 PM, Magnus Fromreide <ma...@lysator.liu.se>wrote:
> On Wed, 2010-10-13 at 09:58 +0000, bvass...@users.sourceforge.net wrote:
> > Revision: 19421
> >
> http://net-snmp.svn.sourceforge.net/net-snmp/?rev=19421&view=rev
> > Author: bvassche
> > Date: 2010-10-13 09:58:34 +0000 (Wed, 13 Oct 2010)
> >
> > Log Message:
> > -----------
> > CHANGES: Cygwin: Running the unit tests does now work (cd testing &&
> ./RUNFULLTESTS -g unit-tests).
> > CHANGES: MinGW: Running the unit tests does now work (cd testing &&
> ./RUNFULLTESTS -g unit-tests).
>
> Some of the changes here do look a tad odd.
>
> -cat > $2.c << EOF1
> +rm -f "$2.c"
> +cat >>"$2.c" <<EOF
>
> Why is rm -f x ; foo >> x better than foo > x?
>
That is the recommended way for creating files from shell scripts when e.g.
a soft link or named pipe could be in the way. The actual reason why I made
that change is because that makes the script more symmetric.
> -#include "$1"
> +EOF
> +cat >>"$2.c" "$1"
> +cat >>"$2.c" <<EOF
>
> This one is interesting - I think the reason for the #include as opposed
> to outright inclusion is to get the error messages reported in the
> correct file and not in some temporary file. Did that not work as
> expected?
>
The #include directive didn't work on one of the platforms I tested, that's
why I replaced it by a "cat".
And from r19423 on, the compiler will again report roper file names and line
numbers when compiling a unit test.
Bart.
------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3.
Spend less time writing and rewriting code and more time creating great
experiences on the web. Be a part of the beta today.
http://p.sf.net/sfu/beautyoftheweb
_______________________________________________
Net-snmp-coders mailing list
Net-snmp-coders@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/net-snmp-coders
