On 15 April 2015 at 20:33, Ivan Zhakov <i...@visualsvn.com> wrote: > On 6 March 2015 at 15:02, <rhuij...@apache.org> wrote: >> >> Author: rhuijben >> Date: Fri Mar 6 12:02:30 2015 >> New Revision: 1664596 >> >> URL: http://svn.apache.org/r1664596 >> Log: >> * subversion/libsvn_subr/config_win.c >> (svn_config__parse_registry): Hide the ENOENT errors from the chain, >> but keep others. >> >> Modified: >> subversion/trunk/subversion/libsvn_subr/config_win.c >> >> Modified: subversion/trunk/subversion/libsvn_subr/config_win.c >> URL: >> http://svn.apache.org/viewvc/subversion/trunk/subversion/libsvn_subr/config_win.c?rev=1664596&r1=1664595&r2=1664596&view=diff >> ============================================================================== >> --- subversion/trunk/subversion/libsvn_subr/config_win.c (original) >> +++ subversion/trunk/subversion/libsvn_subr/config_win.c Fri Mar 6 12:02:30 >> 2015 >> @@ -198,10 +198,12 @@ svn_config__parse_registry(svn_config_t >> if (err != ERROR_SUCCESS) >> { >> apr_status_t apr_err = APR_FROM_OS_ERROR(err); >> + svn_boolean_t is_enoent = APR_STATUS_IS_ENOENT(apr_err); >> >> - if (must_exist || !APR_STATUS_IS_ENOENT(apr_err)) >> + if (must_exist || !is_enoent) >> return svn_error_createf(SVN_ERR_BAD_FILENAME, >> - svn_error_wrap_apr(apr_err, NULL), >> + is_enoent ? NULL >> + : svn_error_wrap_apr(apr_err, >> NULL), >> _("Can't open registry key '%s'"), >> svn_dirent_local_style(file, pool)); > Hi Bert, > > What do you think about refactoring condition above like in attached patch? > I've committed my patch in r1675604.
-- Ivan Zhakov
