moveCamera with CameraUpdateFactory.newLatLngBound

Question

I'm making use of the new Android Google Maps API.

I create an activity which includes a MapFragment. In the activity onResume I set the markers into the GoogleMap object and then define a bounding box for the map which includes all of the markers.

This is using the following pseudo code:

LatLngBounds.Builder builder = new LatLngBounds.Builder();
while(data) {
   LatLng latlng = getPosition();
   builder.include(latlng);
}
CameraUpdate cameraUpdate = CameraUpdateFactory
   .newLatLngBounds(builder.build(), 10);
map.moveCamera(cameraUpdate);

The call to map.moveCamera() causes my application to crash with the following stack:

Caused by: java.lang.IllegalStateException: 
    Map size should not be 0. Most likely, layout has not yet 

    at maps.am.r.b(Unknown Source)
    at maps.y.q.a(Unknown Source)
    at maps.y.au.a(Unknown Source)
    at maps.y.ae.moveCamera(Unknown Source)
    at com.google.android.gms.maps.internal.IGoogleMapDelegate$Stub
        .onTransact(IGoogleMapDelegate.java:83)
    at android.os.Binder.transact(Binder.java:310)
    at com.google.android.gms.maps.internal.IGoogleMapDelegate$a$a
        .moveCamera(Unknown Source)
    at com.google.android.gms.maps.GoogleMap.moveCamera(Unknown Source)
    at ShowMapActivity.drawMapMarkers(ShowMapActivity.java:91)
    at ShowMapActivity.onResume(ShowMapActivity.java:58)
    at android.app.Instrumentation
        .callActivityOnResume(Instrumentation.java:1185)
    at android.app.Activity.performResume(Activity.java:5182)
    at android.app.ActivityThread
        .performResumeActivity(ActivityThread.java:2732)

If - instead of the newLatLngBounds() factory method I use newLatLngZoom() method then the same trap does not occur.

Is the onResume the best place to draw the markers onto the GoogleMap object or should I be drawing the markers and setting the camera position somewhere else?

Answer 1

Another approach would something like (Assuming your topmost view is a FrameLayout named rootContainer, even though it will work as long as you always choose your topmost container no matter which type or name it has):

((FrameLayout)findViewById(R.id.rootContainer)).getViewTreeObserver()
    .addOnGlobalLayoutListener(new OnGlobalLayoutListener() {
        public void onGlobalLayout() {
            layoutDone = true;
        }
    });

Modifying your camera functions to only work if layoutDone is true will solve all your problems without having to add extra functions or wire up logic to the layoutListener handler.

Answer 2

OK I worked this out. As documented here that API can't be used pre-layout.

The correct API to use is described as:

Note: Only use the simpler method newLatLngBounds(boundary, padding) to generate a CameraUpdate if it is going to be used to move the camera after the map has undergone layout. During layout, the API calculates the display boundaries of the map which are needed to correctly project the bounding box. In comparison, you can use the CameraUpdate returned by the more complex method newLatLngBounds(boundary, width, height, padding) at any time, even before the map has undergone layout, because the API calculates the display boundaries from the arguments that you pass.

To fix the problem I calculated my screen size and provided the width and height to

public static CameraUpdate newLatLngBounds(
    LatLngBounds bounds, int width, int height, int padding)

This then allowed me to specify the bounding box pre-layout.

Answer 3

You can use simple newLatLngBounds method in OnCameraChangeListener. All will be working perfectly and you don't need to calculate screen size. This event occurs after map size calculation (as I understand).

Example:

map.setOnCameraChangeListener(new OnCameraChangeListener() {

    @Override
    public void onCameraChange(CameraPosition arg0) {
        // Move camera.
        map.moveCamera(CameraUpdateFactory.newLatLngBounds(builder.build(), 10));
        // Remove listener to prevent position reset on camera move.
        map.setOnCameraChangeListener(null);
    }
});

Answer 4

Ok I'm facing same issue. I have my fragment with my SupportmapFragment, ABS, and navigation drawer. What I did was:

public void resetCamera() {

    LatLngBounds.Builder builderOfBounds = new LatLngBounds.Builder();
    // Set boundaries ...
    LatLngBounds bounds = builderOfBounds.build();
    CameraUpdate cu;
    try{
        cu = CameraUpdateFactory.newLatLngBounds(bounds,10);
        // This line will cause the exception first times 
        // when map is still not "inflated"
        map.animateCamera(cu); 
        System.out.println("Set with padding");
    } catch(IllegalStateException e) {
        e.printStackTrace();
        cu = CameraUpdateFactory.newLatLngBounds(bounds,400,400,0);
        map.animateCamera(cu);
        System.out.println("Set with wh");
    }

    //do the rest...
}

And, by the way, I'm calling resetCamera() from onCreateView after inflating and before returning.
What this does is catch the exception first time (while map "gets a size" as a way of saying it...) and then, other times I need to reset the camera, map already has size and does it through padding.

Issue is explained in documentation, it says:

Do not change the camera with this camera update until the map has undergone layout (in order for this method to correctly determine the appropriate bounding box and zoom level, the map must have a size). Otherwise an IllegalStateException will be thrown. It is NOT sufficient for the map to be available (i.e. getMap() returns a non-null object); the view containing the map must have also undergone layout such that its dimensions have been determined. If you cannot be sure that this has occured, use newLatLngBounds(LatLngBounds, int, int, int) instead and provide the dimensions of the map manually.

I think it's a pretty decent solution. Hope it helps someone.

Answer 5

I used different approach which works for recent versions of Google Maps SDK (9.6+) and based on onCameraIdleListener. As I see so far it's callback method onCameraIdle called always after onMapReady. So my approach looks like this piece of code (considering it put in Activity):

@Override
protected void onCreate(Bundle savedInstanceState) {
    super.onCreate(savedInstanceState);
    // set content view and call getMapAsync() on MapFragment
}

@Override
public void onMapReady(GoogleMap googleMap) {
    map = googleMap;
    map.setOnCameraIdleListener(this);
    // other initialization stuff
}

@Override
public void onCameraIdle() {
    /* 
       Here camera is ready and you can operate with it. 
       you can use 2 approaches here:

      1. Update the map with data you need to display and then set
         map.setOnCameraIdleListener(null) to ensure that further events
         will not call unnecessary callback again.

      2. Use local boolean variable which indicates that content on map
         should be updated
    */
}

Answer 6

Adding and removing markers can be done pre-layout completion, but moving the camera cannot (except using newLatLngBounds(boundary, padding) as noted in the OP's answer).

Probably the best place to perform an initial camera update is using a one-shot OnGlobalLayoutListener as shown in Google's sample code, e.g. see the following excerpt from setUpMap() in MarkerDemoActivity.java:

// Pan to see all markers in view.
// Cannot zoom to bounds until the map has a size.
final View mapView = getSupportFragmentManager()
    .findFragmentById(R.id.map).getView();
if (mapView.getViewTreeObserver().isAlive()) {
    mapView.getViewTreeObserver().addOnGlobalLayoutListener(
    new OnGlobalLayoutListener() {
        @SuppressLint("NewApi") // We check which build version we are using.
        @Override
        public void onGlobalLayout() {
            LatLngBounds bounds = new LatLngBounds.Builder()
                    .include(PERTH)
                    .include(SYDNEY)
                    .include(ADELAIDE)
                    .include(BRISBANE)
                    .include(MELBOURNE)
                    .build();
            if (Build.VERSION.SDK_INT < Build.VERSION_CODES.JELLY_BEAN) {
              mapView.getViewTreeObserver().removeGlobalOnLayoutListener(this);
            } else {
              mapView.getViewTreeObserver().removeOnGlobalLayoutListener(this);
            }
            mMap.moveCamera(CameraUpdateFactory.newLatLngBounds(bounds, 50));
        }
    });
}