feat: Refactor engine interface to use BaseBoard and enhance legal moves retrieval in server

Author: miskibin

draughts/engine.py

@@ -5,6 +5,7 @@
 from loguru import logger
 import numpy as np
 
+from draughts.boards.base import BaseBoard
 from draughts.boards.standard import Board, Move, Figure
 from draughts.models import Color
 
@@ -63,7 +64,7 @@ class Engine(ABC):
 
     @abstractmethod
     def get_best_move(
-        self, board: Board, with_evaluation: bool
+        self, board: BaseBoard, with_evaluation: bool
     ) -> Move | tuple[Move, float]:
         """
         Returns best move for given board.
@@ -138,7 +139,7 @@ def _get_piece_index(self, piece):
         # Map piece value to 0-4 index
         return piece + 2
 
-    def compute_hash(self, board: Board) -> int:
+    def compute_hash(self, board: BaseBoard) -> int:
         h = 0
         for i, piece in enumerate(board._pos):
             if piece != 0:
@@ -147,7 +148,7 @@ def compute_hash(self, board: Board) -> int:
             h ^= self.zobrist_turn
         return h
 
-    def evaluate(self, board: Board) -> float:
+    def evaluate(self, board: BaseBoard) -> float:
         """
         Evaluation function with material and PST.
         Returns score from the perspective of the side to move.
@@ -181,7 +182,7 @@ def evaluate(self, board: Board) -> float:
             return -score
         return score
 
-    def get_best_move(self, board: Board, with_evaluation: bool = False) -> Move | tuple[Move, float]:
+    def get_best_move(self, board: BaseBoard, with_evaluation: bool = False) -> Move | tuple[Move, float]:
         self.start_time = time.time()
         self.nodes = 0
         self.stop_search = False
@@ -240,7 +241,7 @@ def get_best_move(self, board: Board, with_evaluation: bool = False) -> Move | t
             return best_move, float(best_score)
         return best_move
 
-    def negamax(self, board: Board, depth: int, alpha: float, beta: float, h: int) -> float:
+    def negamax(self, board: BaseBoard, depth: int, alpha: float, beta: float, h: int) -> float:
         self.nodes += 1
 
         # Check time
@@ -337,7 +338,7 @@ def negamax(self, board: Board, depth: int, alpha: float, beta: float, h: int) -
 
         return best_value
 
-    def quiescence_search(self, board: Board, alpha: float, beta: float, h: int, qs_depth: int = 0) -> float:
+    def quiescence_search(self, board: BaseBoard, alpha: float, beta: float, h: int, qs_depth: int = 0) -> float:
         """Search captures until position is quiet."""
         self.nodes += 1
 
@@ -377,7 +378,7 @@ def quiescence_search(self, board: Board, alpha: float, beta: float, h: int, qs_
 
         return alpha
 
-    def _update_hash(self, current_hash: int, board: Board, move: Move) -> int:
+    def _update_hash(self, current_hash: int, board: BaseBoard, move: Move) -> int:
         # XOR out source
         start_sq = move.square_list[0]
         piece = board._pos[start_sq]
@@ -405,7 +406,7 @@ def _update_hash(self, current_hash: int, board: Board, move: Move) -> int:
 
         return current_hash
 
-    def _order_moves(self, moves: List[Move], board: Board | None = None, h: int = 0, depth: int = 0) -> List[Move]:
+    def _order_moves(self, moves: List[Move], board: BaseBoard | None = None, h: int = 0, depth: int = 0) -> List[Move]:
         # 1. PV Move from TT
         tt_entry = self.tt.get(h)
         pv_move = tt_entry[3] if tt_entry else None
@@ -438,7 +439,7 @@ def score_move(move):
         moves.sort(key=score_move, reverse=True)
         return moves
 
-    def _order_captures(self, moves: List[Move], board: Board) -> List[Move]:
+    def _order_captures(self, moves: List[Move], board: BaseBoard) -> List[Move]:
         # Sort by number of captures
         moves.sort(key=lambda m: len(m.captured_list), reverse=True)
         return moves

draughts/server/server.py

@@ -187,7 +187,7 @@ def get_position(self, request: Request) -> PositionResponse:
     def get_legal_moves(self) -> dict:
         """Get all legal moves for the current position."""
         with self._lock:
-            moves_dict = defaultdict(list)
+            moves_dict: dict[int, list[int]] = defaultdict(list)
             for move in list(self.board.legal_moves):
                 moves_dict[int(move.square_list[0])].extend(
                     map(int, move.square_list[1:])
@@ -235,7 +235,8 @@ def get_best_move(self, request: Request) -> PositionResponse:
             if not legal_moves:
                 return self.position_json
 
-            move = engine.get_best_move(self.board)
+            result = engine.get_best_move(self.board, with_evaluation=False)
+            move = result if not isinstance(result, tuple) else result[0]
 
             # Validate move is legal (handles stale TT or overlapping requests)
             if move not in legal_moves: