On Thu, May 24, 2012 at 11:23:22AM +0200, Ingo Brückl wrote:
> Reimar Döffinger wrote on Tue, 22 May 2012 20:06:38 +0200:
>
> > On Tue, May 22, 2012 at 11:11:13AM +0200, Ingo Brückl wrote:
> >> Reimar Döffinger wrote on Tue, 22 May 2012 09:14:02 +0200:
> >>
> >> > If someone has a good idea for testing all these headers without having
> >> > to compile for each language it might be worth implementing it.
> >>
> >> Here you go.
>
> > I don't think that actually detects that issue.
>
> It does.
Well, at least with some compilers it prints a warning.
It is not an error though and compiler-specific.
> +export CC
I think other scripts take the compiler as a argument.
> +all: help/help_checked $(ALL_PRG-yes)
It's really testing code, I don't think it should be run by
default.
Because otherwise I'll have to point out everything that can
be a potential issue on any imaginable system.
If it's a separate target I'll be happy with something that
will work on most developer's systems.
> +help/help_checked: help/help_mp*.h
> + help/help_check.sh $?
Which would also mean you could get rid of the
help_checked file and just use a PHONY target.
Actually I guess it wouldn't hurt much to do that
either way.
> + sed -n "s:^#define \([0-9A-Za-z_]\+\) .*:strcmp(\1, \1);:p" "$h" >>
> ${CHECK}.c
I admit I have no idea about what :p means or whether it is
required by POSIX sed. strdup might be a tiny bit simpler.
> + $CC -Werror -c ${CHECK}.c
I think it would be better to use a -o option instead of assuming the
result will be ${CHECK}.o
> + rm -f ${CHECK}.c ${CHECK}.o
Might be better to do as an exit hook, so even on CTRL+C they will be
removed.
_______________________________________________
MPlayer-DOCS mailing list
[email protected]
https://lists.mplayerhq.hu/mailman/listinfo/mplayer-docs
