痛点场景：绘图应用的操作管理

假设你在开发一个绘图App，需要支持：

添加/删除图形
修改图形属性
撤销/重做操作
批量执行命令

传统实现方式：

void _handleAddShape(ShapeType type) {final shape = _createShape(type);setState(() => _shapes.add(shape));
}void _handleDeleteShape(Shape shape) {setState(() => _shapes.remove(shape));
}void _handleChangeColor(Shape shape, Color newColor) {setState(() => shape.color = newColor);
}// 撤销操作？需要自己维护状态...

问题爆发点：

🔄 撤销/重做功能难以实现
📦 批量操作无法封装
🔗 操作与执行代码紧密耦合
⏱️ 延迟执行或排队操作困难

命令模式解决方案

核心思想： 将请求封装为对象，从而允许：

参数化客户端不同请求
排队或记录请求
支持可撤销操作

四个关键角色：

命令接口(Command)： 声明执行操作
具体命令(ConcreteCommand)： 绑定接收者与动作
接收者(Receiver)： 知道如何执行操作
调用者(Invoker)： 触发命令执行

Flutter绘图命令实现

1. 定义命令接口

abstract class DrawingCommand {void execute();void undo();String get description; // 用于命令历史显示
}

2. 实现具体命令

// 添加图形命令
class AddShapeCommand implements DrawingCommand {final List<Shape> _shapes;final Shape _shape;@overrideString get description => '添加 ${_shape.type}';AddShapeCommand(this._shapes, this._shape);@overridevoid execute() => _shapes.add(_shape);@overridevoid undo() => _shapes.remove(_shape);
}// 修改颜色命令
class ChangeColorCommand implements DrawingCommand {final Shape _shape;final Color _newColor;Color _previousColor;@overrideString get description => '修改颜色';ChangeColorCommand(this._shape, this._newColor);@overridevoid execute() {_previousColor = _shape.color;_shape.color = _newColor;}@overridevoid undo() {_shape.color = _previousColor;}
}// 删除图形命令
class DeleteShapeCommand implements DrawingCommand {final List<Shape> _shapes;final Shape _shape;int _index = -1;@overrideString get description => '删除 ${_shape.type}';DeleteShapeCommand(this._shapes, this._shape);@overridevoid execute() {_index = _shapes.indexOf(_shape);_shapes.remove(_shape);}@overridevoid undo() {if (_index != -1) {_shapes.insert(_index, _shape);}}
}

3. 创建命令管理器

class CommandManager {final List<DrawingCommand> _commandHistory = [];final List<DrawingCommand> _undoStack = [];void executeCommand(DrawingCommand command) {command.execute();_commandHistory.add(command);_undoStack.clear();notifyListeners();}void undo() {if (_commandHistory.isEmpty) return;final command = _commandHistory.removeLast();command.undo();_undoStack.add(command);notifyListeners();}void redo() {if (_undoStack.isEmpty) return;final command = _undoStack.removeLast();command.execute();_commandHistory.add(command);notifyListeners();}bool get canUndo => _commandHistory.isNotEmpty;bool get canRedo => _undoStack.isNotEmpty;// 与ChangeNotifier结合final _changeNotifier = ChangeNotifier();void addListener(VoidCallback listener) => _changeNotifier.addListener(listener);void removeListener(VoidCallback listener) => _changeNotifier.removeListener(listener);void notifyListeners() => _changeNotifier.notifyListeners();
}

4. 在Flutter中使用

class DrawingApp extends StatefulWidget {@override_DrawingAppState createState() => _DrawingAppState();
}class _DrawingAppState extends State<DrawingApp> {final List<Shape> _shapes = [];final CommandManager _commandManager = CommandManager();@overridevoid initState() {super.initState();_commandManager.addListener(_refresh);}void _refresh() => setState(() {});void _addCircle() {_commandManager.executeCommand(AddShapeCommand(_shapes, Circle(Colors.blue)),);}void _changeColor(Shape shape) {_commandManager.executeCommand(ChangeColorCommand(shape, Colors.red),);}@overrideWidget build(BuildContext context) {return Scaffold(appBar: AppBar(title: Text('绘图应用'),actions: [IconButton(icon: Icon(Icons.undo),onPressed: _commandManager.canUndo ? _commandManager.undo : null,),IconButton(icon: Icon(Icons.redo),onPressed: _commandManager.canRedo ? _commandManager.redo : null,),],),body: Stack(children: [..._shapes.map((shape) => DraggableShape(shape: shape,onColorChange: () => _changeColor(shape),)),],),floatingActionButton: FloatingActionButton(onPressed: _addCircle,child: Icon(Icons.add),),);}
}

Flutter中的实际应用场景

场景1：宏命令（批量操作）

class MacroCommand implements DrawingCommand {final List<DrawingCommand> _commands = [];void addCommand(DrawingCommand command) {_commands.add(command);}@overrideString get description => '批量操作 (${_commands.length}个命令)';@overridevoid execute() {for (final command in _commands) {command.execute();}}@overridevoid undo() {for (var i = _commands.length - 1; i >= 0; i--) {_commands[i].undo();}}
}// 使用
final macro = MacroCommand()..addCommand(AddShapeCommand(_shapes, Circle(Colors.red)))..addCommand(AddShapeCommand(_shapes, Rectangle(Colors.blue)))..addCommand(ChangeColorCommand(_shapes[0], Colors.green));_commandManager.executeCommand(macro);

场景2：事务操作

class TransactionCommand implements DrawingCommand {final List<DrawingCommand> _commands = [];bool _executed = false;void addCommand(DrawingCommand command) {if (_executed) throw StateError('事务已执行');_commands.add(command);}@overrideString get description => '事务操作';@overridevoid execute() {try {for (final command in _commands) {command.execute();}_executed = true;} catch (e) {// 任何一个命令失败就回滚for (var i = _commands.length - 1; i >= 0; i--) {_commands[i].undo();}rethrow;}}@overridevoid undo() {if (!_executed) return;for (var i = _commands.length - 1; i >= 0; i--) {_commands[i].undo();}_executed = false;}
}

场景3：远程控制（跨平台命令）

abstract class RemoteCommand {Future<void> execute();Map<String, dynamic> toJson();factory RemoteCommand.fromJson(Map<String, dynamic> json) {// 根据json创建具体命令}
}class SaveDrawingCommand implements RemoteCommand {final List<Shape> shapes;@overrideFuture<void> execute() async {await Api.saveDrawing(shapes);}@overrideMap<String, dynamic> toJson() => {'type': 'save','shapes': shapes.map((s) => s.toJson()).toList(),};
}// 通过平台通道执行
MethodChannel('commands').setMethodCallHandler((call) async {final command = RemoteCommand.fromJson(call.arguments);await command.execute();
});

命令模式与状态管理结合

将命令历史与Provider结合：

class CommandHistoryProvider extends ChangeNotifier {final CommandManager _manager;CommandHistoryProvider(this._manager) {_manager.addListener(notifyListeners);}List<String> get commandHistory => _manager._commandHistory.map((c) => c.description).toList();List<String> get undoStack => _manager._undoStack.map((c) => c.description).toList();
}// 在UI中显示历史
Consumer<CommandHistoryProvider>(builder: (context, provider, child) {return Column(children: [Text('操作历史:'),...provider.commandHistory.map((desc) => Text(desc)),SizedBox(height: 20),Text('可重做操作:'),...provider.undoStack.map((desc) => Text(desc)),],);}
)

命令模式最佳实践

何时使用命令模式：
- 需要实现撤销/重做功能
- 需要支持事务操作
- 需要将操作排队或延迟执行
- 需要支持宏命令（命令组合）
- 需要支持跨平台操作

Flutter特化技巧：

// 将命令与Shortcuts绑定
Shortcuts(shortcuts: {LogicalKeySet(LogicalKeyboardKey.control, LogicalKeyboardKey.z): const UndoIntent(),LogicalKeySet(LogicalKeyboardKey.control, LogicalKeyboardKey.y): const RedoIntent(),},child: Actions(actions: {UndoIntent: CallbackAction(onInvoke: (_) => _commandManager.undo(),),RedoIntent: CallbackAction(onInvoke: (_) => _commandManager.redo(),),},child: Builder(builder: (context) => ...),),
)

性能优化：

// 懒执行命令
class LazyCommand implements DrawingCommand {final Future<void> Function() _action;bool _executed = false;@overridevoid execute() async {if (!_executed) {await _action();_executed = true;}}
}

测试策略：

test('撤销应恢复原始状态', () {final shapes = [Circle(Colors.red)];final command = ChangeColorCommand(shapes[0], Colors.blue);command.execute();expect(shapes[0].color, Colors.blue);command.undo();expect(shapes[0].color, Colors.red);
});

命令模式 vs 策略模式

特性	命令模式	策略模式
目的	封装操作请求	封装算法
关注点	何时/如何执行操作	如何完成特定任务
典型应用	撤销/重做/事务	算法替换/策略切换
执行时机	可延迟/排队执行	通常立即执行

命令模式的高级变体

1. 可逆命令工厂

class CommandFactory {static DrawingCommand createAddCommand(List<Shape> shapes, ShapeType type) {return AddShapeCommand(shapes, _createShape(type));}static DrawingCommand createDeleteCommand(List<Shape> shapes, Shape shape) {return DeleteShapeCommand(shapes, shape);}// 注册自定义命令static void register(String type, DrawingCommand Function() creator) {_customCommands[type] = creator;}
}

2. 命令持久化

class PersistentCommand implements DrawingCommand {final SharedPreferences _prefs;final String _key;@overridevoid execute() async {await _prefs.setString(_key, 'executed');}@overridevoid undo() async {await _prefs.remove(_key);}@overrideFuture<bool> get isExecuted async {return _prefs.containsKey(_key);}
}

3. 时间旅行调试

class TimeTravelManager {final List<List<DrawingCommand>> _timeline = [];int _currentState = -1;void snapshot(List<DrawingCommand> commands) {// 移除当前状态之后的所有状态if (_currentState < _timeline.length - 1) {_timeline.removeRange(_currentState + 1, _timeline.length);}_timeline.add(List.from(commands));_currentState = _timeline.length - 1;}void goToState(int index) {if (index >= 0 && index < _timeline.length) {_currentState = index;// 重新执行到目标状态的所有命令}}
}