Re: svn commit: rev 879 - trunk/subversion/include trunk/subversion/libsvn_wc trunk/subversion/libsvn_subr trunk/subversion/tests/libsvn_subr

On Mon, Jan 14, 2002 at 02:31:51PM -0600, cmpilato@collab.net wrote:
> Greg Stein <gstein@lyra.org> writes:
>
> > > - if ((style == svn_wc__eol_style_none)
> > > - && (! revision) && (! author) && (! date) && (! url))
> > > + if ((style == svn_wc__eol_style_none) && (! keywords))
> >
> > What about the case where keywords is supplied, but the pointers are all
> > NULL? Is an app supposed to first check the keywords structure before
> > passing it, decide there isn't anything in it, and pass NULL instead? I'd
> > say "no" :-) ... which means you could get all NULLs and not want to do any
> > translation.
>
> If you get all NULLs, you will, yes, be doing translation, but no
> keywords will be expanded/contracted. So, you *effectively* are doing
> no translation.

Right; that's what I was thinking, so I was hoping that the actual
translation step could be avoided. i.e. optimize partway with the !keywords
as you have, and another step with the check for individual NULLs. (to
hopefully avoid that timestamp change)

Cheers,
-g

-- 
Greg Stein, http://www.lyra.org/
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org