机器人项目部分数据本地化——工厂模式+策略模式实现

场景：

机器人设备上展示的首页数据、其它模块数据均通过后台配置，机器人端获取数据保存，下次判断相关数据有无从本地获取，机器人设备端只负责展示并执行点击事件

弊端：

机器人环境下特别刚开机对网络存在需求，弱网情况下经常性数据无法获取，机器无法正常运行【各种展会弱网情况明显】

解决方案：

所有数据采用本地配置化。

具体方案：

本地数据以Json数据存储在.json文件中，读取后转换成JavaBean的集合即可。【此种方案最简便，且避免了String溢出现象】

具体导入实现只需要一行代码即可：

        InitTableFactory.loadPropertyPath("homepage.json");

  {
    "Md5": "500eafc04370a29b4626433430c1e423",
    "apkName": "null",
    "apkUrl": "null",
    "apkVersion": "null",
    "bak1": "",
    "bak2": "",
    "catagoryId": "aab6438c84494dc1a2a116236df87812",
    "catagoryName": "首页",
    "channel": "儿童早教",
    "channelMark": "儿童早教",
    "id": 19,
    "imageUrl": "file:///android_asset/recomend_configpic/home_children_education.png",
    "type": "2",
    "mark": "",
    "third": "house_children_education",
    "pageUrl": "",
    "no": 2,
    "isTitle": false
  },
  {
    "Md5": "6f1e7e159208d19df1371546debba696",
    "apkName": "null",
    "apkUrl": "null",
    "apkVersion": "null",
    "bak1": "",
    "bak2": "",
    "catagoryId": "aab6438c84494dc1a2a116236df87812",
    "catagoryName": "首页",
    "channel": "音乐",
    "channelMark": "音乐",
    "id": 21,
    "imageUrl": "file:///android_asset/recomend_configpic/home_musice.png",
    "type": "2",
    "mark": "",
    "third": "house_music",
    "pageUrl": "",
    "no": 4,
    "isTitle": false
  },
  {
    "Md5": "e3a82245856bfecfae8848a53aa37881",
    "apkName": "null",
    "apkUrl": "null",
    "apkVersion": "null",
    "bak1": "",
    "bak2": "",
    "catagoryId": "aab6438c84494dc1a2a116236df87812",
    "catagoryName": "首页",
    "channel": "聊天互动",
    "channelMark": "聊天互动",
    "id": 23,
    "imageUrl": "file:///android_asset/recomend_configpic/home_chat.png",
    "type": "2",
    "mark": "",
    "third": "house_chat_interaction",
    "pageUrl": "",
    "no": 6,
    "isTitle": false
  },
  {
    "Md5": "29b681c27b30f56a3f48bd8b8fbe0b0f",
    "apkName": "com.fgecctv.beautycamera",
    "apkUrl": "file:///android_asset/recomend_configpic/0a937326a4d64c97b8b14c08b4f120a5.apk",
    "apkVersion": "1",
    "bak1": "meiyan_apk.apk",
    "bak2": "",
    "catagoryId": "aab6438c84494dc1a2a116236df87812",
    "catagoryName": "首页",
    "channel": "相机",
    "channelMark": "相机",
    "id": 24,
    "imageUrl": "file:///android_asset/recomend_configpic/home_camera.png",
    "type": "0",
    "mark": "美颜apk",
    "third": "xmly",
    "pageUrl": "",
    "no": 7,
    "isTitle": false
  },
  {
    "Md5": "772c573b5b2e4bea21562ab6072b50fb",
    "apkName": "null",
    "apkUrl": "null",
    "apkVersion": "null",
    "bak1": "",
    "bak2": "",
    "catagoryId": "aab6438c84494dc1a2a116236df87812",
    "catagoryName": "首页",
    "channel": "萌相册",
    "channelMark": "萌相册",
    "id": 25,
    "imageUrl": "file:///android_asset/recomend_configpic/home_moe_photoalbum.png",
    "type": "2",
    "mark": "",
    "third": "house_photo_album",
    "pageUrl": "",
    "no": 8,
    "isTitle": false
  }]

具体实现方式：策略模式+工厂模式

工厂类： InitTableFactory
策略接口：TableStrategy
工厂实现：RecommendTable

实现方式可扩展、低耦合。

工厂模式、策略模式复习：

1. 简单工厂模式
   简单工厂模式就是通过调用静态方法来返回（得到）对象的一种思路，这样在需要创建多个对象时，避免了需要先多次创建工厂对象，缺点是某种程度上违反了开放——封闭原则，因为每当需要增加新的对象类型时，就要修改已有的静态工厂类。简单工厂模式（包括别的工厂模式）更侧重于 “得到对象”，一般设计的时候这个对象应该是现实世界中某种事物的映射，有它自己的属性与方法，。

2. 策略模式
   策略模式更偏重于算法实现，按《大话设计模式》中的说法，策略模式更多地使用接口（Interface）而不是父类来实现，（当然这并不是绝对的，对于静态工厂模式中得到的对象的抽象也可以用接口，策略模式也可以用基类。），而接口更侧重于对于行为的抽象而不是对于对象的抽象。策略模式往往侧重于实现方便地替换不同的算法类，而这些类是通过方法来实现某些业务功能的，自身很可能没有属于自己的属性

具体实现如下：

public interface TableStrategy {
     void initTable();
     void load(String path);
}

public class InitTableFactory {


    public static String TAG = "InitTableFactory";

    public static void loadPropertyPath(String path) {
        String fileName = FileUtils.getFileName(path);
        TableStrategy strategy = null;
        switch (fileName) {
            case "homepage.json":
                strategy = new RecommendTable();
                break;
        }
        strategy.load(path);
    }

}

public class RecommendTable  implements TableStrategy {

    @Override
    public void load(String path) {
        List<RecommendConfigInfo> list = SQLite.select().from(RecommendConfigInfo.class).queryList();
         if(list.size()>0)
            return ;
        InputStream dataInputStream = getClass().getResourceAsStream("/assets/"+path);
        try {
            final List<RecommendConfigInfo> recommendConfigInfos= readJsonStream(dataInputStream);
             synchronized (MagicApp.getContext()) {
                FlowManager.getDatabase(RobotDatabase.class).executeTransaction(new ITransaction() {
                    @Override
                    public void execute(DatabaseWrapper databaseWrapper) {
                        if (recommendConfigInfos.size() != 0) {
                            for (RecommendConfigInfo recommendConfigInfo : recommendConfigInfos) {
                                  recommendConfigInfo.insert();
                             }
                        }
                    }
                });
            }

        } catch (IOException e) {
            e.printStackTrace();
        }
    }


    @Override
    public void initTable() {

    }

    public List<RecommendConfigInfo> readJsonStream(InputStream in) throws IOException {
        JsonReader reader = new JsonReader(new InputStreamReader(in, "UTF-8"));
        List<RecommendConfigInfo> list = new ArrayList<>();
        reader.beginArray();
        while (reader.hasNext()) {
              Gson gson = new Gson();
            RecommendConfigInfo message = gson.fromJson(reader, RecommendConfigInfo.class);
            list.add(message);
        }
        reader.endArray();
        reader.close();
        return list;
    }
}

具体实现：RecommendTable 类中 load()方法和readJsonStream