On Mon, 17 Feb 2003, I wrote:
> Thank you for the report. Yes, I can reproduce that effect, and it
> appears to be a bug. I will investigate as soon as I can, and post a
> patch if it is a simple fix.
I regret to announce that the bug was introduced in version 4.11 when I
refactored that code. Unfortunately, there wasn't a test for the final
default of *@ in my test suite. There is now. The simple patch below
fixes the problem.
--
Philip Hazel University of Cambridge Computing Service,
ph10@??? Cambridge, England. Phone: +44 1223 334714.
*** exim-4.12/src/search.c Wed Dec 18 10:28:04 2002
--- search.c Mon Feb 17 14:25:59 2003
***************
*** 748,754 ****
try that. If we do match, the first variable (the wild part) is the whole key,
and the second is empty. */
! if (yield == NULL && (starflags & SEARCH_STAR) != 0)
{
DEBUG(D_lookup) debug_printf("trying to match *\n");
yield = internal_search_find(handle, filename, US"*");
--- 748,754 ----
try that. If we do match, the first variable (the wild part) is the whole key,
and the second is empty. */
! if (yield == NULL && (starflags & (SEARCH_STAR|SEARCH_STARAT)) != 0)
{
DEBUG(D_lookup) debug_printf("trying to match *\n");
yield = internal_search_find(handle, filename, US"*");
