2026-rff_mp/skorohodovsa/task_2/source/strategy/astar.py

import heapq
from typing import Optional

from source.models.base import Cell, Maze
from source.strategy.algorithms import PathFindingStrategy


def _manhattan(a: Cell, b: Cell) -> int:
    """Манхэттенское расстояние между двумя клетками."""
    return abs(a.x - b.x) + abs(a.y - b.y)


class AStarStrategy(PathFindingStrategy):
    """Алгоритм A* с манхэттенской эвристикой."""

    def find_path(self, maze: Maze, start: Optional[Cell] = None, exit: Optional[Cell] = None) -> list[Cell]:
        if start is None:
            start = maze.start
        if exit is None:
            exit = self.exit

        g_score: dict[Cell, int] = {start: 0}
        came_from: dict[Cell, Optional[Cell]] = {start: None}

        counter = 0
        open_heap: list[tuple[int, int, Cell]] = [
            (_manhattan(start, exit), counter, start)
        ]

        while open_heap:
            _, _, current = heapq.heappop(open_heap)

            if current is exit:
                return self._reconstruct_path(came_from, exit)

            for neighbor in maze.get_neighbors(current.x, current.y):
                tentative_g = g_score[current] + 1

                if tentative_g < g_score.get(neighbor, float("inf")):
                    g_score[neighbor] = tentative_g
                    came_from[neighbor] = current
                    f = tentative_g + _manhattan(neighbor, exit)
                    counter += 1
                    heapq.heappush(open_heap, (f, counter, neighbor))

        return []