Joseph Dane <jdane@studio3511.com> writes:
> anyhow, the message I'm using here is long enough to be eye-catching.
> using a printf with an explicit message seems anti-i18n, but that's
> the way the other messages in this file are produced, so ...
Thanks for the patch, it looks good.
I figured out that it was against subversion/clients/cmdline/status.c,
of course. But in general it's helpful to give the long paths for
affected files, in the log message and in the Index lines of the
patch, then there's no need to guess. (There are four `status.c's in
Subversion, and I have all of them in editor buffers at the moment!
:-) )
I committed it in revision 821. Also fixed the documentation for
svn_wc_status_none in wc.h, since it's clear that svn_wc_status_none
is not just for properties. (And I tweaked the actual message status
prints out in this case, to be a little more attention-getting.)
If you have any ambition to write a regression test for this, that
would be wonderful, but no big deal if you haven't sufficient time.
-Karl
> Fix issue #598
>
> * status.c (svn_cl__print_status_list): Display a meaningful message
> when calling 'status' on a non-existant (as opposed to simply
> non-VC'd) file.
>
> Index: ./status.c
> ===================================================================
> --- ./status.c
> +++ ./status.c Thu Jan 10 16:15:21 2002
> @@ -216,10 +216,15 @@
> if ((skip_unrecognized) && (! status->entry))
> continue;
> - if (detailed)
> - print_long_format (path, status);
> - else
> - print_short_format (path, status);
> + if (status->text_status == svn_wc_status_none)
> + printf ("this file does not exist: %s\n", path);
> + else
> + {
> + if (detailed)
> + print_long_format (path, status);
> + else
> + print_short_format (path, status);
> + }
> }
>
>
> --
>
> joe
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Oct 21 14:36:56 2006