对于 Android Developer 来讲,不少开源库都是属于开发必备的知识点,从使用方式到实现原理再到源码解析,这些都须要咱们有必定程度的了解和运用能力。因此我打算来写一系列关于开源库源码解析和实战演练的文章,初定的目标是 EventBus、ARouter、LeakCanary、Retrofit、Glide、OkHttp、Coil 等七个知名开源库,但愿对你有所帮助 😇😇java
公众号:字节数组git
系列文章导航:github
咱们知道,EventBus 在发送了消息后,就会直接回调该消息类型的全部监听方法,回调操做是经过反射 method.invoke
来实现的,那么在回调以前也必须先拿到应用内全部的监听方法才行。EventBus 获取监听方法的方式有两种:数组
这里先介绍第一种方式,这种方式只须要导入以下依赖便可缓存
implementation "org.greenrobot:eventbus:3.2.0"
复制代码
EventBus 经过 EventBus.register(Object)
方法来进行注册的。该方法会对 subscriber 进行解析,经过 SubscriberMethodFinder 的 findSubscriberMethods
方法将 subscriber 包含的全部声明了@Subscribe
注解的方法的签名信息保存到内存中,当有消息被 Post 时就能够直接在内存中查找到目标方法了markdown
public void register(Object subscriber) {
Class<?> subscriberClass = subscriber.getClass();
List<SubscriberMethod> subscriberMethods = subscriberMethodFinder.findSubscriberMethods(subscriberClass);
synchronized (this) {
for (SubscriberMethod subscriberMethod : subscriberMethods) {
subscribe(subscriber, subscriberMethod);
}
}
}
复制代码
从 SubscriberMethod 包含的全部参数能够看出来,它包含了 @Subscribe
的参数信息以及对应的方法签名信息app
public class SubscriberMethod {
final Method method;
final ThreadMode threadMode;
//消息类型
final Class<?> eventType;
//消息优先级
final int priority;
//是否属于黏性消息
final boolean sticky;
/** Used for efficient comparison */
String methodString;
···
}
复制代码
SubscriberMethodFinder 会将每次的查找结果缓存到 METHOD_CACHE
中,这对某些会前后经历屡次注册和反注册操做的 subscriber 来讲比较有用,由于每次查找可能须要依靠屡次循环遍历和反射操做,会稍微有点消耗性能,但缓存也会占用一部份内存空间框架
private static final Map<Class<?>, List<SubscriberMethod>> METHOD_CACHE = new ConcurrentHashMap<>();
List<SubscriberMethod> findSubscriberMethods(Class<?> subscriberClass) {
List<SubscriberMethod> subscriberMethods = METHOD_CACHE.get(subscriberClass);
if (subscriberMethods != null) {
return subscriberMethods;
}
if (ignoreGeneratedIndex) {
subscriberMethods = findUsingReflection(subscriberClass);
} else {
subscriberMethods = findUsingInfo(subscriberClass);
}
if (subscriberMethods.isEmpty()) {
//若是为空,说明没找到使用 @Subscribe 方法,那么 register 操做就是没有意义的,直接抛出异常
throw new EventBusException("Subscriber " + subscriberClass
+ " and its super classes have no public methods with the @Subscribe annotation");
} else {
METHOD_CACHE.put(subscriberClass, subscriberMethods);
return subscriberMethods;
}
}
复制代码
由于ignoreGeneratedIndex
默认是 false,因此这里直接看 findUsingInfo(subscriberClass)
方法async
其主要逻辑是:ide
prepareFindState()
方法从对象池 FIND_STATE_POOL
中获取空闲的 FindState 对象,若是不存在则初始化一个新的,并在使用事后经过 getMethodsAndRelease
方法将对象还给对象池。经过对象池来避免无限制地建立 FindState
对象,这也算作是一个优化点。FindState 用于在反射遍历过程当中保存各类中间状态值findState.subscriberInfo
和 subscriberInfoIndexes
默认都是等于 null,因此主要看 findUsingReflectionInSingleClass
方法便可,从该方法名可知是经过反射操做来进行解析的,解析结果会被存到 findState
中findState.moveToSuperclass()
方法将下一个查找的 class 对象指向父类private static final int POOL_SIZE = 4;
private static final FindState[] FIND_STATE_POOL = new FindState[POOL_SIZE];
private List<SubscriberMethod> findUsingInfo(Class<?> subscriberClass) {
//步骤1
FindState findState = prepareFindState();
findState.initForSubscriber(subscriberClass);
while (findState.clazz != null) {
findState.subscriberInfo = getSubscriberInfo(findState);
if (findState.subscriberInfo != null) {
SubscriberMethod[] array = findState.subscriberInfo.getSubscriberMethods();
for (SubscriberMethod subscriberMethod : array) {
if (findState.checkAdd(subscriberMethod.method, subscriberMethod.eventType)) {
findState.subscriberMethods.add(subscriberMethod);
}
}
} else {
//步骤2
findUsingReflectionInSingleClass(findState);
}
//步骤3
findState.moveToSuperclass();
}
return getMethodsAndRelease(findState);
}
private List<SubscriberMethod> getMethodsAndRelease(FindState findState) {
List<SubscriberMethod> subscriberMethods = new ArrayList<>(findState.subscriberMethods);
findState.recycle();
synchronized (FIND_STATE_POOL) {
//回收 findState,尝试将之存到对象池中
for (int i = 0; i < POOL_SIZE; i++) {
if (FIND_STATE_POOL[i] == null) {
FIND_STATE_POOL[i] = findState;
break;
}
}
}
return subscriberMethods;
}
//若是对象池中有可用的对象则取出来使用,不然的话就构建一个新的
private FindState prepareFindState() {
synchronized (FIND_STATE_POOL) {
for (int i = 0; i < POOL_SIZE; i++) {
FindState state = FIND_STATE_POOL[i];
if (state != null) {
FIND_STATE_POOL[i] = null;
return state;
}
}
}
return new FindState();
}
复制代码
这里来主要看下 findUsingReflectionInSingleClass
方法是如何完成反射操做的。若是解析到的方法签名不符合要求,则在开启了严格检查的状况下直接抛出异常;若是方法签名符合要求,则会将方法签名保存到subscriberMethods
中
private void findUsingReflectionInSingleClass(FindState findState) {
Method[] methods;
try {
// This is faster than getMethods, especially when subscribers are fat classes like Activities
//获取 clazz 包含的全部方法,不包含继承得来的方法
methods = findState.clazz.getDeclaredMethods();
} catch (Throwable th) {
// Workaround for java.lang.NoClassDefFoundError, see https://github.com/greenrobot/EventBus/issues/149
try {
//获取 clazz 以及其父类的全部 public 方法
methods = findState.clazz.getMethods();
} catch (LinkageError error) { // super class of NoClassDefFoundError to be a bit more broad...
String msg = "Could not inspect methods of " + findState.clazz.getName();
if (ignoreGeneratedIndex) {
msg += ". Please consider using EventBus annotation processor to avoid reflection.";
} else {
msg += ". Please make this class visible to EventBus annotation processor to avoid reflection.";
}
throw new EventBusException(msg, error);
}
//因为 getDeclaredMethods() 都抛出异常了,就再也不继续向下循环了,因此指定下次循环时忽略父类
findState.skipSuperClasses = true;
}
for (Method method : methods) {
int modifiers = method.getModifiers();
if ((modifiers & Modifier.PUBLIC) != 0 && (modifiers & MODIFIERS_IGNORE) == 0) {
//method 是 public 的,且不是 ABSTRACT、STATIC、BRIDGE、SYNTHETIC
Class<?>[] parameterTypes = method.getParameterTypes();
if (parameterTypes.length == 1) { //方法包含的参数个数是一
Subscribe subscribeAnnotation = method.getAnnotation(Subscribe.class);
if (subscribeAnnotation != null) { //方法签名包含 Subscribe 注解
Class<?> eventType = parameterTypes[0];
if (findState.checkAdd(method, eventType)) {
//校验经过后,就将 Subscribe 注解的配置信息及 method 方法签名保存起来
ThreadMode threadMode = subscribeAnnotation.threadMode();
findState.subscriberMethods.add(new SubscriberMethod(method, eventType, threadMode,
subscribeAnnotation.priority(), subscribeAnnotation.sticky()));
}
}
} else if (strictMethodVerification && method.isAnnotationPresent(Subscribe.class)) {
//由于 EventBus 只支持包含一个入参参数的注解函数,因此若是开启了严格的方法校验那么就抛出异常
String methodName = method.getDeclaringClass().getName() + "." + method.getName();
throw new EventBusException("@Subscribe method " + methodName +
"must have exactly 1 parameter but has " + parameterTypes.length);
}
} else if (strictMethodVerification && method.isAnnotationPresent(Subscribe.class)) {
//若是 method 的方法签名不符合要求且开启了严格的方法校验那么就抛出异常
String methodName = method.getDeclaringClass().getName() + "." + method.getName();
throw new EventBusException(methodName +
" is a illegal @Subscribe method: must be public, non-static, and non-abstract");
}
}
}
复制代码
findUsingReflectionInSingleClass
方法的一个重点是 findState.checkAdd
方法。若是往简单了想,只要把 subscriber 每一个声明了 Subscribe 注解的方法都给保存起来就能够了,但是还须要考虑一些特殊状况:
checkAdd
方法就用于进行上述判断
//以 eventType 做为 key,method 或者 FindState 做为 value
final Map<Class, Object> anyMethodByEventType = new HashMap<>();
//以 methodKey 做为 key,methodClass 做为 value
final Map<String, Class> subscriberClassByMethodKey = new HashMap<>();
boolean checkAdd(Method method, Class<?> eventType) {
// 2 level check: 1st level with event type only (fast), 2nd level with complete signature when required.
// Usually a subscriber doesn't have methods listening to the same event type.
Object existing = anyMethodByEventType.put(eventType, method);
if (existing == null) {
//existing 等于 null 说明以前未解析到监听相同事件的方法,检查经过
//由于大部分状况下监听者不会声明多个监听相同事件的方法,因此先进行这步检查效率上会比较高
return true;
} else { //existing 不等于 null 说明以前已经解析到一样监听这个事件的方法了
if (existing instanceof Method) {
if (!checkAddWithMethodSignature((Method) existing, eventType)) {
// Paranoia check
throw new IllegalStateException();
}
// Put any non-Method object to "consume" the existing Method
//会执行到这里,说明存在多个方法监听同个 Event,那么将将 eventType 对应的 value 置为 this
//避免屡次检查,让其直接去执行 checkAddWithMethodSignature 方法
anyMethodByEventType.put(eventType, this);
}
return checkAddWithMethodSignature(method, eventType);
}
}
private boolean checkAddWithMethodSignature(Method method, Class<?> eventType) {
methodKeyBuilder.setLength(0);
methodKeyBuilder.append(method.getName());
methodKeyBuilder.append('>').append(eventType.getName());
//以 methodName>eventTypeName 字符串做为 key
//经过这个 key 来判断是否存在子类重写了父类方法的状况
String methodKey = methodKeyBuilder.toString();
//获取声明了 method 的类对应的 class 对象
Class<?> methodClass = method.getDeclaringClass();
Class<?> methodClassOld = subscriberClassByMethodKey.put(methodKey, methodClass);
//1. 若是 methodClassOld == null 为 true,说明 method 是第一次解析到,容许添加
//2. 若是 methodClassOld.isAssignableFrom(methodClass) 为 true
//2.一、说明 methodClassOld 是 methodClass 的父类,须要以子类重写的方法 method 为准,容许添加
// 实际上应该不存在这种状况,由于 EventBus 是从子类开始向父类进行遍历的
//2.二、说明 methodClassOld 是 methodClass 是同个类,即 methodClass 声明了多个方法对同个事件进行监听 ,也容许添加
if (methodClassOld == null || methodClassOld.isAssignableFrom(methodClass)) {
// Only add if not already found in a sub class
return true;
} else {
// Revert the put, old class is further down the class hierarchy
//因为 EventBus 是从子类向父类进行解析
//会执行到这里就说明以前已经解析到了相同 key 的方法,对应子类重写了父类方法的状况
//此时须要以子类重写的方法 method 为准,因此又将 methodClassOld 从新设回去
subscriberClassByMethodKey.put(methodKey, methodClassOld);
return false;
}
}
复制代码
进行上述操做后,就找到了 subscriber 包含的全部监听方法了,这些方法都会保存到 List<SubscriberMethod>
中。拿到全部方法后,register
方法就须要对 subscriber 及其全部监听方法进行归类了
归类的目的是既是为了方便后续操做也是为了提升效率。 由于可能同时存在多个 subscriber 声明了多个对同种类型消息的监听方法,那么就须要将每种消息类型和其当前的全部监听方法对应起来,提升消息的发送效率。并且在 subscriber 解除注册时,也须要将 subscriber 包含的全部监听方法都给移除掉,那么也须要预先进行归类。监听方法也能够设定本身对消息处理的优先级顺序,因此须要预先对监听方法进行排序
public void register(Object subscriber) {
Class<?> subscriberClass = subscriber.getClass();
List<SubscriberMethod> subscriberMethods = subscriberMethodFinder.findSubscriberMethods(subscriberClass);
synchronized (this) {
for (SubscriberMethod subscriberMethod : subscriberMethods) {
subscribe(subscriber, subscriberMethod);
}
}
}
private final Map<Class<?>, CopyOnWriteArrayList<Subscription>> subscriptionsByEventType;
private final Map<Object, List<Class<?>>> typesBySubscriber;
// Must be called in synchronized block
private void subscribe(Object subscriber, SubscriberMethod subscriberMethod) {
Class<?> eventType = subscriberMethod.eventType;
Subscription newSubscription = new Subscription(subscriber, subscriberMethod);
//subscriptionsByEventType 以消息类型 eventType 做为 key,value 存储了全部对该 eventType 的订阅者,提升后续在发送消息时的效率
CopyOnWriteArrayList<Subscription> subscriptions = subscriptionsByEventType.get(eventType);
if (subscriptions == null) {
subscriptions = new CopyOnWriteArrayList<>();
subscriptionsByEventType.put(eventType, subscriptions);
} else {
if (subscriptions.contains(newSubscription)) {
//说明某个 Subscriber 重复注册了
throw new EventBusException("Subscriber " + subscriber.getClass() + " already registered to event "
+ eventType);
}
}
//将订阅者根据消息优先级高低进行排序
int size = subscriptions.size();
for (int i = 0; i <= size; i++) {
if (i == size || subscriberMethod.priority > subscriptions.get(i).subscriberMethod.priority) {
subscriptions.add(i, newSubscription);
break;
}
}
//typesBySubscriber 以订阅者 subscriber 做为 key,value 存储了其订阅的全部 eventType
//用于向外提供某个类是否已注册的功能,也方便后续在 unregister 时移除 subscriber 下的全部监听方法
List<Class<?>> subscribedEvents = typesBySubscriber.get(subscriber);
if (subscribedEvents == null) {
subscribedEvents = new ArrayList<>();
typesBySubscriber.put(subscriber, subscribedEvents);
}
subscribedEvents.add(eventType);
//下面是关于粘性事件的处理,后续再进行介绍
if (subscriberMethod.sticky) {
if (eventInheritance) {
// Existing sticky events of all subclasses of eventType have to be considered.
// Note: Iterating over all events may be inefficient with lots of sticky events,
// thus data structure should be changed to allow a more efficient lookup
// (e.g. an additional map storing sub classes of super classes: Class -> List<Class>).
Set<Map.Entry<Class<?>, Object>> entries = stickyEvents.entrySet();
for (Map.Entry<Class<?>, Object> entry : entries) {
Class<?> candidateEventType = entry.getKey();
if (eventType.isAssignableFrom(candidateEventType)) {
Object stickyEvent = entry.getValue();
checkPostStickyEventToSubscription(newSubscription, stickyEvent);
}
}
} else {
Object stickyEvent = stickyEvents.get(eventType);
checkPostStickyEventToSubscription(newSubscription, stickyEvent);
}
}
}
复制代码
在介绍消息的具体发送步骤前,先来了解下 EventBus 几种不一样的消息执行策略。执行策略由枚举 ThreadMode 来定义,在 @Subscribe
注解中进行声明。执行策略决定了消息接收方是在哪个线程接收到消息的
ThreadMode | 执行线程 | |
---|---|---|
POSTING | 在发送事件的线程中执行 | 直接调用消息接收方 |
MAIN | 在主线程中执行 | 若是事件就是在主线程发送的,则直接调用消息接收方,不然经过 mainThreadPoster 进行处理 |
MAIN_ORDERED | 在主线程中按顺序执行 | 经过 mainThreadPoster 进行处理,以此保证消息处理的有序性 |
BACKGROUND | 在后台线程中按顺序执行 | 若是事件是在主线程发送的,则提交给 backgroundPoster 处理,不然直接调用消息接收方 |
ASYNC | 提交给空闲的后台线程执行 | 将消息提交到 asyncPoster 进行处理 |
执行策略的具体细分逻辑是在 EventBus 类的postToSubscription
方法完成的
private void postToSubscription(Subscription subscription, Object event, boolean isMainThread) {
switch (subscription.subscriberMethod.threadMode) {
case POSTING:
invokeSubscriber(subscription, event);
break;
case MAIN:
if (isMainThread) {
invokeSubscriber(subscription, event);
} else {
mainThreadPoster.enqueue(subscription, event);
}
break;
case MAIN_ORDERED:
if (mainThreadPoster != null) {
mainThreadPoster.enqueue(subscription, event);
} else {
// temporary: technically not correct as poster not decoupled from subscriber
invokeSubscriber(subscription, event);
}
break;
case BACKGROUND:
if (isMainThread) {
backgroundPoster.enqueue(subscription, event);
} else {
invokeSubscriber(subscription, event);
}
break;
case ASYNC:
asyncPoster.enqueue(subscription, event);
break;
default:
throw new IllegalStateException("Unknown thread mode: " + subscription.subscriberMethod.threadMode);
}
}
复制代码
例如,对于 AsyncPoster 来讲,其每接收到一个消息,都会直接在 enqueue
方法中将本身(Runnable)提交给线程池进行处理,而使用的线程池默认是 Executors.newCachedThreadPool()
,该线程池每接收到一个任务都会立刻交由线程进行处理,因此 AsyncPoster 并不保证消息处理的有序性,但在消息处理的及时性方面会比较高,且每次提交给 AsyncPoster 的消息可能都是由不一样的线程来处理
class AsyncPoster implements Runnable, Poster {
private final PendingPostQueue queue;
private final EventBus eventBus;
AsyncPoster(EventBus eventBus) {
this.eventBus = eventBus;
queue = new PendingPostQueue();
}
public void enqueue(Subscription subscription, Object event) {
PendingPost pendingPost = PendingPost.obtainPendingPost(subscription, event);
queue.enqueue(pendingPost);
eventBus.getExecutorService().execute(this);
}
@Override
public void run() {
PendingPost pendingPost = queue.poll();
if(pendingPost == null) {
throw new IllegalStateException("No pending post available");
}
eventBus.invokeSubscriber(pendingPost);
}
}
复制代码
而 BackgroundPoster 会将任务依次缓存到 PendingPostQueue 中,每次只取出一个任务交由线程池来执行,因此 BackgroundPoster 会保证消息队列在处理时的有序性,但在消息处理的及时性方面相比 AsyncPoster 要低一些
final class BackgroundPoster implements Runnable, Poster {
private final PendingPostQueue queue;
private final EventBus eventBus;
private volatile boolean executorRunning;
BackgroundPoster(EventBus eventBus) {
this.eventBus = eventBus;
queue = new PendingPostQueue();
}
public void enqueue(Subscription subscription, Object event) {
PendingPost pendingPost = PendingPost.obtainPendingPost(subscription, event);
synchronized (this) {
queue.enqueue(pendingPost);
if (!executorRunning) {
executorRunning = true;
eventBus.getExecutorService().execute(this);
}
}
}
···
}
复制代码
而无论是使用什么消息处理策略,最终都是经过调用如下方法来反射调用监听方法
void invokeSubscriber(PendingPost pendingPost) {
Object event = pendingPost.event;
Subscription subscription = pendingPost.subscription;
PendingPost.releasePendingPost(pendingPost);
if (subscription.active) {
invokeSubscriber(subscription, event);
}
}
void invokeSubscriber(Subscription subscription, Object event) {
try {
subscription.subscriberMethod.method.invoke(subscription.subscriber, event);
} catch (InvocationTargetException e) {
handleSubscriberException(subscription, event, e.getCause());
} catch (IllegalAccessException e) {
throw new IllegalStateException("Unexpected exception", e);
}
}
复制代码
EventBus.getDefault().post(Any)
方法用于发送非黏性消息。EventBus 会经过 ThreadLocal 为每一个发送消息的线程维护一个 PostingThreadState 对象,用于为每一个线程维护一个消息队列及其它辅助参数
/** * For ThreadLocal, much faster to set (and get multiple values). */
final static class PostingThreadState {
final List<Object> eventQueue = new ArrayList<>();
boolean isPosting;
boolean isMainThread;
Subscription subscription;
Object event;
boolean canceled;
}
private final ThreadLocal<PostingThreadState> currentPostingThreadState = new ThreadLocal<PostingThreadState>() {
@Override
protected PostingThreadState initialValue() {
return new PostingThreadState();
}
};
/** * Posts the given event to the event bus. */
public void post(Object event) {
PostingThreadState postingState = currentPostingThreadState.get();
List<Object> eventQueue = postingState.eventQueue;
//将消息添加到消息队列
eventQueue.add(event);
if (!postingState.isPosting) {
//是否在主线程发送的消息
postingState.isMainThread = isMainThread();
//标记当前正在发送消息中
postingState.isPosting = true;
if (postingState.canceled) {
throw new EventBusException("Internal error. Abort state was not reset");
}
try {
while (!eventQueue.isEmpty()) {
postSingleEvent(eventQueue.remove(0), postingState);
}
} finally {
postingState.isPosting = false;
postingState.isMainThread = false;
}
}
}
复制代码
每次 post 进来的消息都会先存到消息队列 eventQueue
中,而后经过 while 循环进行处理,消息处理逻辑是经过 postSingleEvent
方法来完成的
其主要逻辑是:
sendNoSubscriberEvent
为 true,那么就主动发送一个 NoSubscriberEvent 事件,用于向外通知消息没有找到任何接收者cancelEventDelivery
方法来拦截事件,再也不继续向下发送。但只有在 POSTING
模式下才能拦截事件,由于只有在这个模式下才能保证监听方法是按照严格的前后顺序被执行的最终,发送的消息都会经过 postToSubscription
方法来完成,根据接收者方法不一样的处理策略进行处理
private void postSingleEvent(Object event, PostingThreadState postingState) throws Error {
Class<?> eventClass = event.getClass();
//用于标记是否有找到消息的接收者
boolean subscriptionFound = false;
if (eventInheritance) {
//步骤2
List<Class<?>> eventTypes = lookupAllEventTypes(eventClass);
int countTypes = eventTypes.size();
for (int h = 0; h < countTypes; h++) {
Class<?> clazz = eventTypes.get(h);
subscriptionFound |= postSingleEventForEventType(event, postingState, clazz);
}
} else {
//步骤3
subscriptionFound = postSingleEventForEventType(event, postingState, eventClass);
}
if (!subscriptionFound) {
if (logNoSubscriberMessages) {
logger.log(Level.FINE, "No subscribers registered for event " + eventClass);
}
if (sendNoSubscriberEvent && eventClass != NoSubscriberEvent.class &&
eventClass != SubscriberExceptionEvent.class) {
//步骤4
post(new NoSubscriberEvent(this, event));
}
}
}
private boolean postSingleEventForEventType(Object event, PostingThreadState postingState, Class<?> eventClass) {
CopyOnWriteArrayList<Subscription> subscriptions;
synchronized (this) {
//找到全部监听者
subscriptions = subscriptionsByEventType.get(eventClass);
}
if (subscriptions != null && !subscriptions.isEmpty()) {
for (Subscription subscription : subscriptions) {
postingState.event = event;
postingState.subscription = subscription;
boolean aborted;
try {
postToSubscription(subscription, event, postingState.isMainThread);
aborted = postingState.canceled;
} finally {
postingState.event = null;
postingState.subscription = null;
postingState.canceled = false;
}
//步骤5
if (aborted) {
break;
}
}
return true;
}
return false;
}
复制代码
黏性消息的意义是为了使得在消息发出来后,即便是后续再进行 register
的 subscriber 也能够收到以前发送的消息,这须要将 @Subscribe
注解的 sticky
属性设为 true,即代表消息接收方但愿接收黏性消息
EventBus.getDefault().postSticky(Any)
方法就用于发送黏性消息。黏性事件会被保存到 stickyEvents
这个 Map 中,key 是 event 的 Class 对象,value 是 event 自己,这也说明对于同一类型的黏性消息来讲,只会保存其最后一个消息
private final Map<Class<?>, Object> stickyEvents;
/** * Posts the given event to the event bus and holds on to the event (because it is sticky). The most recent sticky * event of an event's type is kept in memory for future access by subscribers using {@link Subscribe#sticky()}. */
public void postSticky(Object event) {
synchronized (stickyEvents) {
stickyEvents.put(event.getClass(), event);
}
// Should be posted after it is putted, in case the subscriber wants to remove immediately
post(event);
}
复制代码
对于一个黏性消息,会有两种不一样的时机被 subscriber 接收到
这里主要看第二种状况。register 操做会在 subscribe 方法里完成黏性事件的分发。和 post 操做同样,发送黏性事件时也须要考虑 event 的继承关系
private void subscribe(Object subscriber, SubscriberMethod subscriberMethod) {
···
if (subscriberMethod.sticky) {
if (eventInheritance) {
// Existing sticky events of all subclasses of eventType have to be considered.
// Note: Iterating over all events may be inefficient with lots of sticky events,
// thus data structure should be changed to allow a more efficient lookup
// (e.g. an additional map storing sub classes of super classes: Class -> List<Class>).
//事件类型须要考虑其继承关系
//所以须要判断每个 stickyEvent 的父类型是否存在监听者,有的话就须要都进行回调
Set<Map.Entry<Class<?>, Object>> entries = stickyEvents.entrySet();
for (Map.Entry<Class<?>, Object> entry : entries) {
Class<?> candidateEventType = entry.getKey();
if (eventType.isAssignableFrom(candidateEventType)) {
Object stickyEvent = entry.getValue();
checkPostStickyEventToSubscription(newSubscription, stickyEvent);
}
}
} else {
//事件类型不须要考虑其继承关系
Object stickyEvent = stickyEvents.get(eventType);
checkPostStickyEventToSubscription(newSubscription, stickyEvent);
}
}
}
private void checkPostStickyEventToSubscription(Subscription newSubscription, Object stickyEvent) {
if (stickyEvent != null) {
// If the subscriber is trying to abort the event, it will fail (event is not tracked in posting state)
// --> Strange corner case, which we don't take care of here.
postToSubscription(newSubscription, stickyEvent, isMainThread());
}
}
复制代码
移除指定的黏性事件能够经过如下方法来实现,都是用于将指定事件从 stickyEvents
中移除
/** * Remove and gets the recent sticky event for the given event type. * * @see #postSticky(Object) */
public <T> T removeStickyEvent(Class<T> eventType) {
synchronized (stickyEvents) {
return eventType.cast(stickyEvents.remove(eventType));
}
}
/** * Removes the sticky event if it equals to the given event. * * @return true if the events matched and the sticky event was removed. */
public boolean removeStickyEvent(Object event) {
synchronized (stickyEvents) {
Class<?> eventType = event.getClass();
Object existingEvent = stickyEvents.get(eventType);
if (event.equals(existingEvent)) {
stickyEvents.remove(eventType);
return true;
} else {
return false;
}
}
}
复制代码
解除注册的目的是为了不内存泄露,EventBus 使用了单例模式,若是不主动解除注册的话,EventBus 就会一直持有 subscriber。解除注册是经过 unregister
方法来实现的,该方法逻辑也比较简单,只是将 subscriber 以及其关联的全部 method 对象从集合中移除而已
而此处虽然会将关于 subscriber 的信息均给移除掉,可是在 SubscriberMethodFinder 中的静态成员变量 METHOD_CACHE
依然会缓存着已经注册过的 subscriber 的信息,这也是为了在某些 subscriber 会前后屡次注册 EventBus 时能够作到信息复用,避免屡次循环反射
/** * Unregisters the given subscriber from all event classes. */
public synchronized void unregister(Object subscriber) {
List<Class<?>> subscribedTypes = typesBySubscriber.get(subscriber);
if (subscribedTypes != null) {
for (Class<?> eventType : subscribedTypes) {
unsubscribeByEventType(subscriber, eventType);
}
typesBySubscriber.remove(subscriber);
} else {
logger.log(Level.WARNING, "Subscriber to unregister was not registered before: " + subscriber.getClass());
}
}
/** * Only updates subscriptionsByEventType, not typesBySubscriber! Caller must update typesBySubscriber. */
private void unsubscribeByEventType(Object subscriber, Class<?> eventType) {
List<Subscription> subscriptions = subscriptionsByEventType.get(eventType);
if (subscriptions != null) {
int size = subscriptions.size();
for (int i = 0; i < size; i++) {
Subscription subscription = subscriptions.get(i);
if (subscription.subscriber == subscriber) {
subscription.active = false;
subscriptions.remove(i);
i--;
size--;
}
}
}
}
复制代码
使用注解处理器(Annotation Processing Tool)能够避免 subscriber 进行注册时的屡次循环反射操做,极大提高了 EventBus 的运行效率。注解处理器是一种注解处理工具,用来在编译期扫描和处理注解,经过注解来生成 Java 文件。即以注解做为桥梁,经过预先规定好的代码生成规则来自动生成 Java 文件。此类注解框架的表明有 ButterKnife、Dragger二、EventBus 等
Java API 已经提供了扫描源码并解析注解的框架,开发者能够经过继承 AbstractProcessor 类来实现本身的注解解析逻辑。APT 的原理就是在注解了某些代码元素(如字段、函数、类等)后,在编译时编译器会检查 AbstractProcessor 的子类,而且自动调用其 process()
方法,而后将添加了指定注解的全部代码元素做为参数传递给该方法,开发者再根据注解元素在编译期输出对应的 Java 代码
关于 APT 技术的原理和应用能够看这篇文章:Android APT 实例讲解
在 Kotlin 环境引入注解处理器的方法以下所示:
apply plugin: 'kotlin-kapt'
kapt {
arguments {
arg('eventBusIndex', 'github.leavesc.demo.MyEventBusIndex')
}
}
dependencies {
implementation "org.greenrobot:eventbus:3.2.0"
kapt "org.greenrobot:eventbus-annotation-processor:3.2.0"
}
复制代码
当中,MyEventBusIndex 就是在编译阶段将生成的辅助文件,github.leavesc.demo.MyEventBusIndex
就是生成的辅助文件的包名路径,能够由咱们本身定义
原始文件:
/** * 做者:leavesC * 时间:2020/10/01 12:17 * 描述: * GitHub:https://github.com/leavesC */
class MainActivity : AppCompatActivity() {
override fun onCreate(savedInstanceState: Bundle?) {
super.onCreate(savedInstanceState)
setContentView(R.layout.activity_main)
}
@Subscribe
fun fun1(msg: String) {
}
@Subscribe(threadMode = ThreadMode.MAIN, priority = 100)
fun fun2(msg: String) {
}
}
复制代码
编译事后生成的辅助文件以下所示。能够看出,MyEventBusIndex 文件中封装了 subscriber 和其全部监听方法的签名信息,这样咱们就无需在运行时再来进行解析了,而是直接在编译阶段就生成好了
/** This class is generated by EventBus, do not edit. */
public class MyEventBusIndex implements SubscriberInfoIndex {
private static final Map<Class<?>, SubscriberInfo> SUBSCRIBER_INDEX;
static {
SUBSCRIBER_INDEX = new HashMap<Class<?>, SubscriberInfo>();
putIndex(new SimpleSubscriberInfo(MainActivity.class, true, new SubscriberMethodInfo[] {
new SubscriberMethodInfo("fun1", String.class),
new SubscriberMethodInfo("fun2", String.class, ThreadMode.MAIN, 100, false),
}));
}
private static void putIndex(SubscriberInfo info) {
SUBSCRIBER_INDEX.put(info.getSubscriberClass(), info);
}
@Override
public SubscriberInfo getSubscriberInfo(Class<?> subscriberClass) {
SubscriberInfo info = SUBSCRIBER_INDEX.get(subscriberClass);
if (info != null) {
return info;
} else {
return null;
}
}
}
复制代码
须要注意的是,在生成了辅助文件后,还须要经过这些类文件来初始化 EventBus
EventBus.builder().addIndex(MyEventBusIndex()).installDefaultEventBus()
复制代码
注入的辅助文件会被保存到 SubscriberMethodFinder 类的成员变量 subscriberInfoIndexes
中,findUsingInfo
方法会先尝试从辅助文件中获取 SubscriberMethod,只有在获取不到的时候才会经过性能较低的反射操做来完成
private List<SubscriberMethod> findUsingInfo(Class<?> subscriberClass) {
FindState findState = prepareFindState();
findState.initForSubscriber(subscriberClass);
while (findState.clazz != null) {
//在没有使用注解处理器的状况下,findState.subscriberInfo 和 subscriberInfoIndexes 的默认值都是为 null,因此 getSubscriberInfo 会返回 null
//此时就须要经过 findUsingReflectionInSingleClass 方法来进行反射获取
//而在有使用注解处理器的状况下,subscriberInfoIndexes 就存储了自动生成的辅助文件,此时 getSubscriberInfo 就能够从辅助文件中拿到目标信息
//从而避免了反射操做
findState.subscriberInfo = getSubscriberInfo(findState);
if (findState.subscriberInfo != null) {
SubscriberMethod[] array = findState.subscriberInfo.getSubscriberMethods();
for (SubscriberMethod subscriberMethod : array) {
if (findState.checkAdd(subscriberMethod.method, subscriberMethod.eventType)) {
findState.subscriberMethods.add(subscriberMethod);
}
}
} else {
findUsingReflectionInSingleClass(findState);
}
findState.moveToSuperclass();
}
return getMethodsAndRelease(findState);
}
private SubscriberInfo getSubscriberInfo(FindState findState) {
if (findState.subscriberInfo != null && findState.subscriberInfo.getSuperSubscriberInfo() != null) {
SubscriberInfo superclassInfo = findState.subscriberInfo.getSuperSubscriberInfo();
if (findState.clazz == superclassInfo.getSubscriberClass()) {
return superclassInfo;
}
}
if (subscriberInfoIndexes != null) {
for (SubscriberInfoIndex index : subscriberInfoIndexes) {
SubscriberInfo info = index.getSubscriberInfo(findState.clazz);
if (info != null) {
return info;
}
}
}
return null;
}
复制代码
使用了注解处理器后也有必定的弊端。因为 MyEventBusIndex 是经过静态常量类型的 Map 来保存全部的方法签名信息,当在初始化 EventBus 时该 Map 就同时被初始化了,这就至关于在一开始就进行了全量加载,而某些 subscriber 咱们可能不会使用到,这就形成了内存浪费。而若是是经过反射来获取,那就至关于在按需加载,只有 subscriber 进行注册了才会去缓存 subscriber 带有的监听方法
上文有介绍到,子类能够继承父类的 Subscribe 方法。但有一个比较奇怪的地方是:若是子类重写了父类多个 Subscribe 方法的话,就会抛出 IllegalStateException。例如,在下面的例子中。父类 BaseActivity 声明了两个 Subscribe 方法,子类 MainActivity 重写了这两个方法,此时运行后就会抛出 IllegalStateException。而若是 MainActivity 不重写或者只重写一个方法的话,就能够正常运行
/** * 做者:leavesC * 时间:2020/10/01 12:49 * 描述: * GitHub:https://github.com/leavesC */
open class BaseActivity : AppCompatActivity() {
@Subscribe
open fun fun1(msg: String) {
}
@Subscribe
open fun fun2(msg: String) {
}
}
class MainActivity : BaseActivity() {
override fun onCreate(savedInstanceState: Bundle?) {
super.onCreate(savedInstanceState)
setContentView(R.layout.activity_main)
EventBus.getDefault().register(this)
}
override fun onDestroy() {
super.onDestroy()
EventBus.getDefault().unregister(this)
}
@Subscribe
override fun fun1(msg: String) {
}
@Subscribe
override fun fun2(msg: String) {
}
}
复制代码
按道理来讲,若是子类重写了父类一个 Subscribe 方法均可以正常使用的话,那么重写两个也应该能够正常使用才对。但是上述例子就表现得 EventBus 好像有 bug 似的。经过定位堆栈信息,能够发现是在 FindState
的 checkAdd
方法抛出了异常
其抛出异常的步骤是这样的:
checkAdd
方法开始解析 BaseActivity
的 fun2
方法时,existing
对象就是 BaseActivity.fun1
,此时就会执行到操做1,而因为子类已经重写了 fun1
方法,此时 checkAddWithMethodSignature
方法就会返回 false,最终致使抛出异常boolean checkAdd(Method method, Class<?> eventType) {
// 2 level check: 1st level with event type only (fast), 2nd level with complete signature when required.
// Usually a subscriber doesn't have methods listening to the same event type.
Object existing = anyMethodByEventType.put(eventType, method);
if (existing == null) {
return true;
} else {
if (existing instanceof Method) {
//操做1
if (!checkAddWithMethodSignature((Method) existing, eventType)) {
// Paranoia check
throw new IllegalStateException();
}
// Put any non-Method object to "consume" the existing Method
anyMethodByEventType.put(eventType, this);
}
return checkAddWithMethodSignature(method, eventType);
}
}
复制代码
EventBus 有一个 issues 也反馈了这个问题:issues,该问题在 2018 年时就已经存在了,EeventBus 的做者也只是回复说:只在子类进行方法监听
removeStickyEvent
方法会有一个比较让人误解的点:对于经过 EventBus.getDefault().postSticky(XXX)
方法发送的黏性消息没法经过 removeStickyEvent
方法来使现有的监听者拦截该事件
例如,假设下面的两个方法都已经处于注册状态了,postSticky
后,即便在 fun1
方法中移除了黏性消息,fun2
方法也能够接收到消息。这是由于 postSticky
方法最终也是要靠调用 post
方法来完成消息发送,而 post
方法并不受 stickyEvents
的影响
@Subscribe(sticky = true)
fun fun1(msg: String) {
EventBus.getDefault().removeStickyEvent(msg)
}
@Subscribe(sticky = true)
fun fun2(msg: String) {
}
复制代码
而若是 EventBus 中已经存储了黏性事件,那么在上述两个方法刚 register 时,fun1
方法就能够拦截住消息使 fun2
方法接收不到消息。这是由于 register
方法是在 for 循环中遍历 method,若是以前的方法已经移除了黏性消息的话,那么后续方法就没有黏性消息须要处理了
public void register(Object subscriber) {
Class<?> subscriberClass = subscriber.getClass();
List<SubscriberMethod> subscriberMethods = subscriberMethodFinder.findSubscriberMethods(subscriberClass);
synchronized (this) {
//在 for 循环中遍历 method
for (SubscriberMethod subscriberMethod : subscriberMethods) {
subscribe(subscriber, subscriberMethod);
}
}
}
复制代码
EventBus 的源码解析到这里就结束了,本文所讲的内容应该也已经涵盖了大部份内容了。这里再来为 EventBus 的实现流程作一个总结
EventBus 的实现思路并不算多难,难的是在实现的时候能够方方面面都考虑周全,作到稳定高效,从 2018 年到 2020 年也才发布了两个版本(也许是做者懒得更新?)原理懂了,那么下一篇就进入实战篇,本身来动手实现一个 EventBus 😇😇