Added debug branch, to make optimized run faster

shenzhen_solitaire/board_actions.py

@@ -16,15 +16,15 @@ class Action:
         pass
 
     def apply(self, action_board: board.Board) -> None:
-        #print(f"Applying {str(self)}")
         assert self._before_state == 0
         assert self._after_state == 0
-        self._before_state = action_board.state_identifier
+        if __debug__:
+            self._before_state = action_board.state_identifier
         self._apply(action_board)
-        self._after_state = action_board.state_identifier
+        if __debug__:
+            self._after_state = action_board.state_identifier
 
     def undo(self, action_board: board.Board) -> None:
-        #print(f"Undoing {str(self)}")
         assert action_board.state_identifier == self._after_state
         self._undo(action_board)
         assert action_board.state_identifier == self._before_state

shenzhen_solitaire/solver.py

@@ -56,6 +56,11 @@ class SolitaireSolver:
 
     def solve(self) -> Iterator[List[board_actions.Action]]:
         while self.stack:
+            if len(self.stack) == -1:
+                self.stack.pop()
+                self.stack.action_stack[-1].undo(self.search_board)
+                assert (self.search_board.state_identifier
+                        in self.state_set)
 
             assert (self.search_board.state_identifier ==
                     self.stack.state_stack[-1])