From: Pasi Kallinen <paxed@alt.org>
Date: Fri, 4 Mar 2022 14:42:11 +0000 (+0200)
Subject: Fix checking monster moving into a trap return value
X-Git-Url: https://granicus.if.org/sourcecode?a=commitdiff_plain;h=2cdd8f8dcbdebf44e4e7e9d552e01d1c5d877629;p=nethack

Fix checking monster moving into a trap return value

When I added a new trap return value, this was one place that
should've checked it.

This was really annoying to debug - it manifested as a temp level
file loading error, because a monster moved on to a level teleport
trap, which zeroes out (mx, my), then later in the monster movement,
a web spinner created a web at (0,0), and then hero leaving the level,
the traps were saved into the level file ... and when loaded, the code
thought a trap with x == 0 meant there were no more traps.
---

diff --git a/src/monmove.c b/src/monmove.c
index f0ead084e..9b5803123 100644
--- a/src/monmove.c
+++ b/src/monmove.c
@@ -1441,6 +1441,8 @@ m_move(register struct monst* mtmp, register int after)
                 didseeit = canseeit;
 
         if (mmoved == MMOVE_MOVED) {
+            int trapret;
+
             /* normal monster move will already have <nix,niy>,
                but pet dog_move() with 'goto postmov' won't */
             nix = mtmp->mx, niy = mtmp->my;
@@ -1475,7 +1477,8 @@ m_move(register struct monst* mtmp, register int after)
             }
 
             newsym(omx, omy); /* update the old position */
-            if (mintrap(mtmp, NO_TRAP_FLAGS) == Trap_Killed_Mon) {
+            trapret = mintrap(mtmp, NO_TRAP_FLAGS);
+            if (trapret == Trap_Killed_Mon || trapret == Trap_Moved_Mon) {
                 if (mtmp->mx)
                     newsym(mtmp->mx, mtmp->my);
                 return 2; /* it died */