Warning this project is not maintained! Check out https://developers.google.com/cloud-messaging/network-manager for a possibly better solution.
A backport of Android Lollipop's JobScheduler to api 10+.
All JobScheduler features are implemented. However, this library has not been well-tested, so I would advise not using in production at this time. There are no guarantees that this will not run down your battery or cause your device to explode.
compile 'me.tatarka.support:jobscheduler:0.1.1'<dependency>
<groupId>me.tatarka.support</groupId>
<artifactId>jobscheduler</artifactId>
<version>0.1.1</version>
<type>aar</type>
</dependency>You also have to enable manifest merging to ensure the services and receivers are added to your manifest.
<plugin>
<groupId>com.jayway.maven.plugins.android.generation2</groupId>
<artifactId>android-maven-plugin</artifactId>
<configuration>
<mergeManifests>true</mergeManifests>
</configuration>
</plugin>The api is identical to the official JobScheduler, the only differences will be what you import.
First create a service to run your job.
import me.tatarka.support.job.JobParameters;
import me.tatarka.support.job.JobService;
/**
* Service to handle callbacks from the JobScheduler. Requests scheduled with the JobScheduler
* ultimately land on this service's "onStartJob" method.
*/
public class TestJobService extends JobService {
@Override
public boolean onStartJob(JobParameters params) {
// Start your job in a seperate thread, calling jobFinished(params, needsRescheudle) when you are done.
// See the javadoc for more detail.
return true;
}
@Override
public boolean onStopJob(JobParameters params) {
// Stop the running job, returing true if it needs to be recheduled.
// See the javadoc for more detail.
return true;
}Then use the JobScheduler to schedule the job.
import me.tatarka.support.job.JobInfo;
import me.tatarka.support.job.JobScheduler;
import me.tatarka.support.os.PersistableBundle;
// Get an instance of the JobScheduler, this will delegate to the system JobScheduler on api 21+
// and to a custom implementataion on older api levels.
JobScheduler jobScheduler = JobScheduler.getInstance(context);
// Extras for your job.
PersistableBundle extras = new PersistableBundle();
extras.putString("key", "value");
// Construct a new job with your service and some constraints.
// See the javadoc for more detail.
JobInfo job = new JobInfo.Builder(0 /*jobid*/, new ComponentName(context, TestJobService.class))
.setMinimumLatency(1000)
.setOverrideDeadline(2000)
.setRequiredNetworkType(JobInfo.NETWORK_TYPE_UNMETERED)
.setRequiresCharging(true)
.setExtras(extras)
.build();
jobScheudler.scheduleJob(job);Finally, register your JobService in the Manifest. Since on api < 21 the
JobScheduler runs within your app, you may need additional permissions
depending on what you are doing.
<?xml version="1.0" encoding="utf-8"?>
<manifest xmlns:android="http://schemas.android.com/apk/res/android"
package="me.tatarka.support.job.sample">
<!-- Always required on api < 21, needed to keep a wake lock while your job is running -->
<uses-permission android:name="android.permission.WAKE_LOCK" />
<!-- Required on api < 21 if you are using setRequiredNetworkType(int) -->
<uses-permission android:name="android.permission.ACCESS_NETWORK_STATE" />
<!-- Required on all api levels if you are using setPersisted(true) -->
<uses-permission android:name="android.permission.RECEIVE_BOOT_COMPLETED" />
<application>
...
<!-- Define your service, make sure to add the permision! -->
<service
android:name=".service.TestJobService"
android:permission="android.permission.BIND_JOB_SERVICE"
android:exported="true" />
</application>
</manifest>-
Pending jobs will be stored in
<privateappdir>/system/job/jobs.xml. Do not delete this file as it will causing jobs to not run. -
If you use
setRequiresDeviceIdle(true)then it may not immediately run in the first idle window. If your app is not running, it will no longer receive notifications on when the device is idle, therefore, it will awake every 91 minutes to check. This is hopefully a reasonable compromise between preserving battery life and ensuring your job is run. -
There may be other subtle differences on when and how many jobs run at the same time. This is because, unlike the system
JobScheduler, the current implementatoin cannot detect the state of other jobs running on the system and batch them together.
The best way to help out right now is by testing, run it for a while and see if it behaves like expected. Creating apps that would help log long-term usage would be helpful as well.
If you want to do something more specific, feel free to leave a comment on any of the unassigned issues, or create your own.
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent