本地广播LocalBroadcastManager
说到组件间通信第一个肯定想到广播BroadcastReceiver,但是这里要说的是一个更优的选择---本地广播LocalBroadcastManager;
-
优点:只在app内传播, 信息不会泄露,也不会被别人的广播干扰, 且比全局广播更高效;
-
缺点:但是本地广播传输消息时将一切都交给系统负责,无法干预传输中的步骤;
-
使用观察者模式
使用demo:
class LocalBroadcastActivity : AppCompatActivity() {
override fun onCreate(savedInstanceState: Bundle?) {
super.onCreate(savedInstanceState)
setContentView(R.layout.activity_local_broadcast)
testLocalBroadcast()
}
private lateinit var broadcastReceiver: BroadcastReceiver
private lateinit var lbm: LocalBroadcastManager
private val localAction = "com.ljy.publicdemo.localAction"
private fun testLocalBroadcast() {
broadcastReceiver = object : BroadcastReceiver() {
/**
* 接收并出列广播
*/
override fun onReceive(context: Context?, intent: Intent?) {
if (localAction == intent!!.action) {
val value = intent.getStringExtra("key_001")
LogUtils.d("key_001=$value, do something... ")
}
}
}
//创建
lbm = LocalBroadcastManager.getInstance(this)
//注册
lbm.registerReceiver(broadcastReceiver, IntentFilter(localAction))
}
private fun sendBroadcast() {
//发送
val intent = Intent(localAction)
intent.putExtra("key_001", "value_001")
lbm.sendBroadcast(intent)
}
fun onBtnClick(view: View) {
when (view.id) {
R.id.button_send_broadcast -> sendBroadcast()
}
}
override fun onDestroy() {
super.onDestroy()
//解绑
lbm.unregisterReceiver(broadcastReceiver)
}
}
本质:看一下LocalBroadcastManager源码
//构造方法如下,其本质还是用handler通信
private LocalBroadcastManager(Context context) {
mAppContext = context;
mHandler = new Handler(context.getMainLooper()) {
@Override
public void handleMessage(Message msg) {
switch (msg.what) {
case MSG_EXEC_PENDING_BROADCASTS:
executePendingBroadcasts();
break;
default:
super.handleMessage(msg);
}
}
};
}
//内部类ReceiverRecord:将receiver和intentFilter封装成ReceiverRecord对象,
private static final class ReceiverRecord {
final IntentFilter filter;
final BroadcastReceiver receiver;
boolean broadcasting;
boolean dead;
ReceiverRecord(IntentFilter _filter, BroadcastReceiver _receiver) {
filter = _filter;
receiver = _receiver;
}
}
//内部类BroadcastRecord:将ReceiverRecord对象封装在BroadcastRecord对象中
private static final class BroadcastRecord {
final Intent intent;
final ArrayList<ReceiverRecord> receivers;
BroadcastRecord(Intent _intent, ArrayList<ReceiverRecord> _receivers) {
intent = _intent;
receivers = _receivers;
}
}
//有三个集合
//一个以receiver为key、以ReceiverRecord列表为value的map
private final HashMap<BroadcastReceiver, ArrayList<ReceiverRecord>> mReceivers= new HashMap<>();
//一个以action为key,以ReceiverRecord列表为value的map
private final HashMap<String, ArrayList<ReceiverRecord>> mActions = new HashMap<>();
//BroadcastRecord对象的集合
private final ArrayList<BroadcastRecord> mPendingBroadcasts = new ArrayList<>();
//注册和解绑的源码就不贴了
//registerReceiver()就是装填mReceivers,mActions这两个map
//unregisterReceiver()就是清除mReceivers,mActions
//发送广播源码
public boolean sendBroadcast(@NonNull Intent intent) {
synchronized (mReceivers) {
//。。。省略部分源码,主要就是对intent中的信息进行校验
//最关键的是下面的装填mPendingBroadcasts,和发送handler消息
if (receivers != null) {
for (int i=0; i<receivers.size(); i++) {
receivers.get(i).broadcasting = false;
}
mPendingBroadcasts.add(new BroadcastRecord(intent, receivers));
if (!mHandler.hasMessages(MSG_EXEC_PENDING_BROADCASTS)) {
mHandler.sendEmptyMessage(MSG_EXEC_PENDING_BROADCASTS);
}
return true;
}
}
}
return false;
}
//还有一个同步发送广播的方法
public void sendBroadcastSync(@NonNull Intent intent) {
if (sendBroadcast(intent)) {
executePendingBroadcasts();
}
}
//上面源码看出一个至关重要的方法executePendingBroadcasts,才是真正调用广播处理的回调
void executePendingBroadcasts() {
while (true) {
final BroadcastRecord[] brs;
synchronized (mReceivers) {
final int N = mPendingBroadcasts.size();
if (N <= 0) {
return;
}
brs = new BroadcastRecord[N];
mPendingBroadcasts.toArray(brs);
mPendingBroadcasts.clear();
}
for (int i=0; i<brs.length; i++) {
final BroadcastRecord br = brs[i];
final int nbr = br.receivers.size();
for (int j=0; j<nbr; j++) {
final ReceiverRecord rec = br.receivers.get(j);
if (!rec.dead) {
rec.receiver.onReceive(mAppContext, br.intent);
}
}
}
}
}
事件总线
- 组件化层级障碍:
由组件化架构图可以看出,组件间相互独立,没有依赖,也就没有关系,也就无法传递信息,那么要如何交流呢?
这时就需要用到基础层base module, 因为组件层的模块都依赖于基础层;
- 事件总线
Android中activity,fragment,service间信息传递相对复杂, 如果用系统级别的广播, 有耗时,容易被捕捉,无法干预,可传输数据类型少等弊端,
于是大佬们就搞出了事件总线;
下面是三种目前常用的事件总线框架:
1. EventBus:
一款针对Android优化的发布/订阅事件总线,主要功能是替代handler,intent,broadcast;
优点是开销小,代码优雅,将发送者和接收者解耦;
/**
* 演示eventBus的使用
*/
//1. 定义要传递的事件实体
data class TestEvent(val params: String)
class EventBusActivity : AppCompatActivity() {
override fun onCreate(savedInstanceState: Bundle?) {
super.onCreate(savedInstanceState)
setContentView(R.layout.activity_eventbus)
}
//2. 准备订阅者
@Subscribe(threadMode = ThreadMode.MAIN)
fun onMessageEvent(event: TestEvent) {
LjyLogUtil.d(event.params)
}
//3. 注册订阅者
override fun onStart() {
super.onStart()
EventBus.getDefault().register(this)
}
//4. 解绑订阅者
override fun onStop() {
super.onStop()
EventBus.getDefault().unregister(this)
}
fun onBtnClick(view: View) {
when (view.id) {
//5. 发送事件, 实际场景中发送事件的发送者一般在其他页面或模块中
R.id.btn_send_event -> EventBus.getDefault().post(TestEvent("测试一下事件发送"))
}
}
}
EventBus2.x的版本和3.x是有很大区别的:
1. 2.x使用的是运行时注解,采用了反射的方式对整个注册的类的所有方法进行扫描来完成注册,因而会对性能有一定影响;
2. 3.x使用的是编译时注解,Java文件会编译成.class文件,再对class文件进行打包等一系列处理。在编译成.class文件时,
EventBus会使用EventBusAnnotationProcessor注解处理器读取@Subscribe()注解并解析、处理其中的信息,
然后生成Java类来保存所有订阅者的订阅信息。这样就创建出了对文件或类的索引关系,并将其编入到apk中。
3. 从EventBus3.0开始使用了对象池缓存减少了创建对象的开销。
EventBus给Android开发者世界带来了一种新的框架和思想,就是消息的发布和订阅。这种思想在其后很多框架中都得到了应用。
2. RxBUs
RxBus不是一个库,而是一个文件,实现只有短短30行代码。RxBus本身不需要过多分析,它的强大完全来自于它基于的RxJava技术。
public class RxBus {
/**
* Instance of {@link Bus}
*/
private static Bus sBus;
/**
* Get the instance of {@link Bus}
*
* @return
*/
public static synchronized Bus get() {
if (sBus == null) {
sBus = new Bus(ThreadEnforcer.ANY);
}
return sBus;
}
}
RxBus有很多实现,如:
AndroidKnife/RxBus(https://github.com/AndroidKnife/RxBus)(即上面的代码)
Blankj/RxBus(https://github.com/Blankj/RxBus)
下面是使用rxbus的demo
//1. rxbus的实现类
package com.ljy.publicdemo.util;
import androidx.lifecycle.Lifecycle;
import androidx.lifecycle.LifecycleOwner;
import com.trello.lifecycle2.android.lifecycle.AndroidLifecycle;
import com.trello.rxlifecycle2.LifecycleProvider;
import java.util.Map;
import java.util.concurrent.ConcurrentHashMap;
import io.reactivex.Observable;
import io.reactivex.ObservableEmitter;
import io.reactivex.ObservableOnSubscribe;
import io.reactivex.subjects.PublishSubject;
import io.reactivex.subjects.Subject;
/**
* @Author: LiuJinYang
* @CreateDate: 2020/5/11 17:01
*/
public class RxBusUtil {
private volatile static RxBusUtil mDefaultInstance;
private final Subject<Object> mBus;
private final Map<Class<?>, Object> mStickyEventMap;
private RxBusUtil() {
mBus = PublishSubject.create().toSerialized();
mStickyEventMap = new ConcurrentHashMap<>();
}
public static RxBusUtil getInstance() {
if (mDefaultInstance == null) {
synchronized (RxBusUtil.class) {
if (mDefaultInstance == null) {
mDefaultInstance = new RxBusUtil();
}
}
}
return mDefaultInstance;
}
/**
* 发送事件
*/
public void post(Object event) {
mBus.onNext(event);
}
/**
* 使用Rxlifecycle解决RxJava引起的内存泄漏
*/
public <T> Observable<T> toObservable(LifecycleOwner owner, final Class<T> eventType) {
LifecycleProvider<Lifecycle.Event> provider = AndroidLifecycle.createLifecycleProvider(owner);
return mBus.ofType(eventType).compose(provider.<T>bindToLifecycle());
}
/**
* 判断是否有订阅者
*/
public boolean hasObservers() {
return mBus.hasObservers();
}
public void reset() {
mDefaultInstance = null;
}
/**
* Stciky 相关
*/
/**
* 发送一个新Sticky事件
*/
public void postSticky(Object event) {
synchronized (mStickyEventMap) {
mStickyEventMap.put(event.getClass(), event);
}
post(event);
}
/**
* 根据传递的 eventType 类型返回特定类型(eventType)的 被观察者
* 使用Rxlifecycle解决RxJava引起的内存泄漏
*/
public <T> Observable<T> toObservableSticky(LifecycleOwner owner, final Class<T> eventType) {
synchronized (mStickyEventMap) {
LifecycleProvider<Lifecycle.Event> provider = AndroidLifecycle.createLifecycleProvider(owner);
Observable<T> observable = mBus.ofType(eventType).compose(provider.<T>bindToLifecycle());
final Object event = mStickyEventMap.get(eventType);
if (event != null) {
return observable.mergeWith(Observable.create(new ObservableOnSubscribe<T>() {
@Override
public void subscribe(ObservableEmitter<T> subscriber) throws Exception {
subscriber.onNext(eventType.cast(event));
}
}));
} else {
return observable;
}
}
}
/**
* 根据eventType获取Sticky事件
*/
public <T> T getStickyEvent(Class<T> eventType) {
synchronized (mStickyEventMap) {
return eventType.cast(mStickyEventMap.get(eventType));
}
}
/**
* 移除指定eventType的Sticky事件
*/
public <T> T removeStickyEvent(Class<T> eventType) {
synchronized (mStickyEventMap) {
return eventType.cast(mStickyEventMap.remove(eventType));
}
}
/**
* 移除所有的Sticky事件
*/
public void removeAllStickyEvents() {
synchronized (mStickyEventMap) {
mStickyEventMap.clear();
}
}
}
//2. 使用
class RxBusActivity:AppCompatActivity(){
override fun onCreate(savedInstanceState: Bundle?) {
super.onCreate(savedInstanceState)
setContentView(R.layout.activity_rx_bus)
registerRxBus()
}
private fun registerRxBus() {
RxBusUtil.getInstance().toObservable(this, TestEvent::class.java).subscribe(object : Observer<TestEvent> {
override fun onSubscribe(d: Disposable) {
LjyLogUtil.d("onSubscribe:回调返回Disposable 对象,让观察者管理订阅状态, 例如取消订阅")
}
override fun onNext(event: TestEvent) {
LjyLogUtil.d("onNext:数据接收处理")
LjyLogUtil.d(event.params)
}
override fun onError(e: Throwable) {
LjyLogUtil.d("onError:发生异常,终止事件流")
}
override fun onComplete() {
LjyLogUtil.d("onComplete:事件流结束")
}
})
}
fun onBtnClick(view: View) {
when(view.id){
R.id.btn_rx_bus->RxBusUtil.getInstance().post(TestEvent("测试rxBus发送消息"))
}
}
}
//因为用了Rxlifecycle,推出activity时无需解绑
与EventBus相比的优点,其实也就是rxJava的优点:
1、RxJava的Observable有onError、onComplete等状态回调。
2、RxJava使用组合而非嵌套的方式,避免了回调地狱。
3、RxJava的线程调度设计的更加优秀,更简单易用。
4、RxJava可使用多种操作符来进行链式调用来实现复杂的逻辑。
5、RxJava的信息效率高于EventBus2.x,低于EventBus3.x。
那么技术选型时如何取舍呢?如果项目中使用了RxJava,则使用RxBus,否则使用EventBus3.x
3. LiveDataBus
LiveDataBus是基于LiveData实现的类似EventBus的消息通信框架,它是基于LiveData实现的,完全可以代替EventBus,RxBus;
为什么会有他呢?
Handler : 容易导致内存泄漏,空指针,高耦合,不利于维护
EventBus :原理实现复杂,无法混淆,需要手动绑定生命周期
RxBus:依赖于RxJava,包太大,影响apk大小,app启动时间
//初代实现如下
ublic final class LiveDataBus {
private final Map<String, MutableLiveData<Object>> bus;
private LiveDataBus() {
bus = new HashMap<>();
}
private static class SingletonHolder {
private static final LiveDataBus DATA_BUS = new LiveDataBus();
}
public static LiveDataBus get() {
return SingletonHolder.DATA_BUS;
}
public <T> MutableLiveData<T> getChannel(String target, Class<T> type) {
if (!bus.containsKey(target)) {
bus.put(target, new MutableLiveData<>());
}
return (MutableLiveData<T>) bus.get(target);
}
public MutableLiveData<Object> getChannel(String target) {
return getChannel(target, Object.class);
}
}
//使用
注册订阅:
LiveDataBus.get().getChannel("key_test", Boolean.class)
.observe(this, new Observer<Boolean>() {
@Override
public void onChanged(@Nullable Boolean aBoolean) {
}
});
发送消息:
LiveDataBus.get().getChannel("key_test").setValue(true);
其项目地址如下:
https://github.com/JinYangLiu/LiveEventBus
另外的项目SmartEventBus,基于LiveEventBus实现,能让你定制自己的消息总线
https://github.com/JinYangLiu/SmartEventBus
4. 组件化事件总线的考量
其实目前常用的各种事件总线xxBus原理都差不多
那么在组件化项目中如何使用这些事件总线呢
1.EventBus,RxBus: 将xxEvent消息容器和事件总线框架的依赖放到base module,其他模块组件依赖于base module;
但是这样每个模块改动都需要增删改baseModule中的消息容器, 组件化要求功能模块独立, 各组件应该尽量避免影响base module;
- LiveDataBus: 无需建立消息模型,但无法想前两者一样拥有类名索引,无法引导正确的编写代码,也无法传递自定义实体到其他模块;
- 组件化中使用EventBus,RxBus,为了更大程度的解耦,可以独立出一个事件总线module,添加事件的实体都在这个module中,
base module依赖 这个事件总线module对事件通信的解耦, 抽离事件到事件总线module中减少对base module的影响;
网友评论