Dean Roehrich <Dean.Roehrich at Sun.COM> writes:
> Well, my team was way behind on cstyle.pl and I finally noticed this...so
> I updated from our October 2005 version to the current version.
>
> This cset:
>
> changeset: 4950:f0c49956a749
> user: Richard Lowe <richlowe at richlowe.net>
> date: Mon Aug 13 20:07:24 2007 -0400
> summary: 306 cdm should whine at people about checks as early as is
> sensible
>
> has the unfortunate side-effect of making cstyle exit after processing one
> file, whether or not there were warnings, and no matter how many files are
> listed on the commandline.
>
> Perhaps the following patch would be appropriate. This removes the exit from
> within the while() loop, and moves it up, after the calls to cstyle().
>
> Dean
>
>
> ------
> Index: webrev/usr/src/tools/scripts/cstyle.pl
> ===================================================================
> --- webrev.orig/usr/src/tools/scripts/cstyle.pl 2008-06-25
> 16:16:11.000000000 -0500
> +++ webrev/usr/src/tools/scripts/cstyle.pl 2008-06-25 16:16:29.917111000
> -0500
> @@ -184,6 +184,7 @@ if ($#ARGV >= 0) {
> } else {
> &cstyle("<stdin>", *STDIN);
> }
> +exit $err_stat;
>
> my $no_errs = 0; # set for CSTYLED-protected lines
>
> @@ -689,7 +690,6 @@ if ($prev eq "") {
> err("last line in file is blank");
> }
>
> -exit $err_stat;
> }
Yes. Can you verify whether jstyle is also a victim?
-- Rich
