JobScheduler和JobService是安卓在api 21中增加的接口，用于在某些指定条件下执行后台任务。
JobScheduler
JobScheduler是用于计划基于应用进程的多种类型任务的api接口。
- 对象获取方法：
[Context.getSystemService(Context.JOB_SCHEDULER_SERVICE)] - 使用
JobInfo.Builder.JobInfo.Builder(int, android.content.ComponentName)构造JobInfo对象，并作为参数传给JobSechduler的schedule(JobInfo)方法。 - 当JobInfo中声明的执行条件满足时，系统会在应用的JobService中启动执行这个任务。
当任务执行时，系统会为你的应用持有WakeLock，所以应用不需要做多余的确保设备唤醒的工作。
JobService
public abstract class JobService
extends Service
JobService继承自Service，是用于处理JobScheduler中规划的异步请求的特殊Service
- 使用JobService必须先在AndroidManifest.xml中声明service和权限
<service android:name="MyJobService" android:permission="android.permission.BIND_JOB_SERVICE"/ > - 应用需要实现onStartJob(JobParameters)接口，在其中执行任务逻辑。
- 这个Service会在一个运行在主线程的Handler中执行规划的任务，所以应用需要在另外的thread/handler/AsyncTask中执行业务逻辑，如果不这么做的话可能会引起主线程的阻塞。
- onStopJob(android.app.job.JobParameters)接口是当计划的执行条件“不再”满足时被执行的（例如网络中断）。
设置周期性任务
private static final int JOB_INFO_ID = 10001;
private static final long JOB_PERIODIC = 5 * 1000L;
private void onJobStartClick() {
JobScheduler jobScheduler = (JobScheduler) getSystemService(JOB_SCHEDULER_SERVICE);
ComponentName componentName = new ComponentName(this, MyJobService.class);
JobInfo jobinfo = new JobInfo.Builder(JOB_INFO_ID, componentName)
.setPeriodic(JOB_PERIODIC)
.build();
}
周期periodic设置为5秒->运行,会发现JobService没有被启动,为什么？
setPeriodic的最小执行间隔
源码里来找答案：android/frameworks/base/core/java/android/app/job/JobInfo.java
/* Minimum interval for a periodic job, in milliseconds. */
private static final long MIN_PERIOD_MILLIS = 15 * 60 * 1000L; // 15 minutes
/* Minimum flex for a periodic job, in milliseconds. */
private static final long MIN_FLEX_MILLIS = 5 * 60 * 1000L; // 5 minutes
/**
* Specify that this job should recur with the provided interval, not more than once per
* period. You have no control over when within this interval this job will be executed,
* only the guarantee that it will be executed at most once within this interval.
* Setting this function on the builder with {@link #setMinimumLatency(long)} or
* {@link #setOverrideDeadline(long)} will result in an error.
* @param intervalMillis Millisecond interval for which this job will repeat.
*/
public Builder setPeriodic(long intervalMillis) {
return setPeriodic(intervalMillis, intervalMillis);
}
/**
* Specify that this job should recur with the provided interval and flex. The job can
* execute at any time in a window of flex length at the end of the period.
* @param intervalMillis Millisecond interval for which this job will repeat. A minimum
* value of {@link #getMinPeriodMillis()} is enforced.
* @param flexMillis Millisecond flex for this job. Flex is clamped to be at least
* {@link #getMinFlexMillis()} or 5 percent of the period, whichever is
* higher.
*/
public Builder setPeriodic(long intervalMillis, long flexMillis) {
mIsPeriodic = true;
mIntervalMillis = intervalMillis;
mFlexMillis = flexMillis;
mHasEarlyConstraint = mHasLateConstraint = true;
return this;
}
/**
* Query the minimum interval allowed for periodic scheduled jobs. Attempting
* to declare a smaller period that this when scheduling a job will result in a
* job that is still periodic, but will run with this effective period.
*
* @return The minimum available interval for scheduling periodic jobs, in milliseconds.
*/
public static final long getMinPeriodMillis() {
return MIN_PERIOD_MILLIS;
}
/**
* Set to the interval between occurrences of this job. This value is <b>not</b> set if the
* job does not recur periodically.
*/
public long getIntervalMillis() {
final long minInterval = getMinPeriodMillis();
return intervalMillis >= minInterval ? intervalMillis : minInterval;
}
总结几个要点：
- 可以看到系统默认设置了一个最小间隔时间15分钟，在获取执行间隔时，会先比较最小间隔时间和设置的间隔时间，取其中大的那个。所以setPeriodic设置时间小于15分钟是不会生效的。
- flexMillis参数是用来设置周期任务执行的活动时间的，这意味着JobScheduler规划的任务不是在精确的时间执行的。并且这个时间也是有最小值的，系统默认5分钟。
- setMinimumLatency和setOverrideDeadline不能同setPeriodic一起使用，会引起报错，另外还有一些其他规则，看源码：
/**
* @return The job object to hand to the JobScheduler. This object is immutable.
*/
public JobInfo build() {
// Allow jobs with no constraints - What am I, a database?
if (!mHasEarlyConstraint && !mHasLateConstraint && mConstraintFlags == 0 &&
mNetworkType == NETWORK_TYPE_NONE &&
mTriggerContentUris == null) {
throw new IllegalArgumentException("You're trying to build a job with no " +
"constraints, this is not allowed.");
}
// Check that a deadline was not set on a periodic job.
if (mIsPeriodic) {
if (mMaxExecutionDelayMillis != 0L) {
throw new IllegalArgumentException("Can't call setOverrideDeadline() on a " +
"periodic job.");
}
if (mMinLatencyMillis != 0L) {
throw new IllegalArgumentException("Can't call setMinimumLatency() on a " +
"periodic job");
}
if (mTriggerContentUris != null) {
throw new IllegalArgumentException("Can't call addTriggerContentUri() on a " +
"periodic job");
}
}
if (mIsPersisted) {
if (mTriggerContentUris != null) {
throw new IllegalArgumentException("Can't call addTriggerContentUri() on a " +
"persisted job");
}
if (!mTransientExtras.isEmpty()) {
throw new IllegalArgumentException("Can't call setTransientExtras() on a " +
"persisted job");
}
if (mClipData != null) {
throw new IllegalArgumentException("Can't call setClipData() on a " +
"persisted job");
}
}
if (mBackoffPolicySet && (mConstraintFlags & CONSTRAINT_FLAG_DEVICE_IDLE) != 0) {
throw new IllegalArgumentException("An idle mode job will not respect any" +
" back-off policy, so calling setBackoffCriteria with" +
" setRequiresDeviceIdle is an error.");
}
JobInfo job = new JobInfo(this);
....
return job;
}
如何查看自己的JobService的运行？
adb给我们提供了dumpsys工具：
adb shell dumpsys jobscheduler
JOB #u0a122/10001: 945a633 com.fantasy.android.demo/.android.job.MyJobService
u0a122 tag=*job*/com.fantasy.android.demo/.android.job.MyJobService
Source: uid=u0a122 user=0 pkg=com.fantasy.android.demo
JobInfo:
Service: com.fantasy.android.demo/.android.job.MyJobService
PERIODIC: interval=+1h0m0s0ms flex=+1h0m0s0ms
Requires: charging=false batteryNotLow=false deviceIdle=false
Backoff: policy=1 initial=+30s0ms
Has early constraint
Has late constraint
Required constraints: TIMING_DELAY DEADLINE
Satisfied constraints: APP_NOT_IDLE DEVICE_NOT_DOZING
Unsatisfied constraints: TIMING_DELAY DEADLINE
Tracking: TIME
Enqueue time: -5m38s906ms
Run time: earliest=+54m21s65ms, latest=+1h54m21s65ms
Last successful run: 2018-01-02 13:07:16
Ready: false (job=false user=true !pending=true !active=true !backingup=true comp=true)
JobService的使用：
继承JobService实现他的两个接口
public class MyJobService extends JobService{
private static final String TAG = "MyJobService";
@Override
public boolean onStartJob(JobParameters jobParameters) {
Log.d(TAG, "onStartJob-->");
return false;
}
@Override
public boolean onStopJob(JobParameters jobParameters) {
Log.d(TAG, "onStopJob-->");
return false;
}
}
Android准备好执行任务时，服务就会启动，此时会在主线程上收到onStartJob()方法调用。
该方法返回false结果表示该任务已经全部做完，此时系统会解绑该JobService，最终会调用JobService的onDestroy()方法，效果就如同自己调用了jobFinished()方法一样。
返回 true 结果则表示任务已经启动成功，但还没有全部做完，此时可以在任务完成后，应用自行调用jobFinished方法。onStopJob(JobParameters)方法是在中断任务时调用，例如用户需要服务在有充电时才运行，如果在调用JobFinished()之前（任务完成之前）充电器拔掉，onStopJob(…) 方法就会被调用，也就是说，一切任务就立即停止了。
调用 onStopJob(…) 方法就是表明，服务马上就要被停掉，这里返回 true 表示：“任务应该计划在下次继续。”返回 false 表示：“不管怎样，事情就到此结束吧，不要计划下次了。”关于jobFinished()
public final void jobFinished (JobParameters params, boolean wantsReschedule)
Job的任务执行完毕后，APP端自己调用，用以通知JobScheduler已经完成了任务。
注意:该方法执行完后不会回调onStopJob(),但是会回调onDestroy()
当wantsReschedule参数设置为true时，表示任务需要另外规划时间进行执行。
而这个执行的时间受限与JobInfo的退避规则。
设置退避规则
setBackoffCriteria
public [JobInfo.Builder] setBackoffCriteria (long initialBackoffMillis, int backoffPolicy)
- initialBackoffMillis是第一次尝试重试的等待间隔，单位为毫秒，预设的参数有：
DEFAULT_INITIAL_BACKOFF_MILLIS
30000
MAX_BACKOFF_DELAY_MILLIS
18000000 - backoffPolicy是对应的退避策略，预设的参数有：
BACKOFF_POLICY_EXPONENTIAL
二进制退避，等待间隔呈指数增长
retry_time(current_time, num_failures) = current_time + initial_backoff_millis * 2 ^ (num_failures - 1), num_failures >= 1
BACKOFF_POLICY_LINEAR
线性退避，等待间隔呈线性增长
retry_time(current_time, num_failures) = current_time + initial_backoff_millis * num_failures, num_failures >= 1
如何查看JobScheduler为应用持有的WakeLock
adb shell dumpsys power com.fantasy.android.demo | grep Wake
参考文章：
https://blog.csdn.net/allisonchen/article/details/79240163
https://blog.csdn.net/allisonchen/article/details/79218713
https://www.cnblogs.com/chase1/p/7221916.html
