When and why deliverResult is called in AsyncTaskLoader?

I want to establish a good understanding of the AsyncTaskLoader lifecycle. I checked several resources, everything is clear but the usage of deliverResult. According to this picture from the internet (available here):

onStartLoading will be called, then if there is data already loaded, deliverResult is called, then it deliver the result to onLoadFinished. However, if there is no data foceLoad will be called, then loadInBackground, then deliverResult, then onLoadFinished.

I did the same way croworc answer suggests here: What does AsyncTaskLoader.deliverResult() actually do?

This is the code:

public class WeatherLoader extends AsyncTaskLoader<List<Weather>> {

    List <Weather> receivedData;
    /** Tag for log messages */


    private String mUrl;


public WeatherLoader(Context context, String url) {
        super(context);
        mUrl = url;
    }

    @Override
    protected void onStartLoading() {
        if (receivedData == null){
            Log.i ("loader ", "No data yet");
            forceLoad();
        } else {
            deliverResult(receivedData);
            Log.i ("loader ", "data is available no reload");
        }

    }

    @Override
    public void deliverResult(List<Weather> data) {
        receivedData = data;

        super.deliverResult(data);
        Log.i ("loader ", "deliver result");
    }

    @Override
    public List<Weather> loadInBackground() {
        Log.i ("loader ", "load in background");
        if (mUrl == null) {
            return null;
        }

        // Perform the network request, parse the response, and extract a list of earthquakes.
        List<Weather> weather = getweatherData(mUrl);
        return weather;
    }

}

But this is the sequence of the callbacks I'm getting when I initialize the loader or restart it: onCreatLoader

No data yet

load in background

onLoaderFinish

deliver result

What really confuses me is that deliverResult is called after onLoaderFinished which also I think contradicts with this page of this book:

available here

The check for the availability of the data used in onStartLoading which calls deliverResult only gets called when the activity is stopped and restarted, like if I navigate to another activity then get back to it. Here is what gets printed in the logcat in this case:

deliver result
data is available no reload

Even onLoadFinished doesn't get called in this case. However, if I do the same behavior of navigating to another activity and getting back to the first one with having deliverResult with its original behavior (where I only call the super version of it), onStart gets called, then loadInBackground, then onLoadFinished, then DeliverResult. So, a new load happens

Can anyone please clarify why this behavior of callbacks is taking place? Does this mean that the image that shows the lifecycle is inaccurate?

Thanks.

Put the log calls before calling super and check the flow sequence again.

onLoadFinished is called during the call to super.deliverResult.

     @Override
    public void deliverResult(List<Weather> data) {

        Log.i ("loader ", "deliver result");
        receivedData = data;

        super.deliverResult(data);

    }