This lesson describes how you can easily take any expensive method and wrap the call inside an
RxJava Observable using the
Observable.defer() and how we can use this to wrap any method in an
Observable so that we can defer the execution of an expensive method until the correct time, and so that we can control which threads are used to execute the method.
We provide an example project where we use
Observable.just() together to execute a slow method on a background thread and return the result on the Android main thread.
We describe why simply using
Observable.just() on its own can't help us solve this problem.
Summary of Content:
- Description of problem; how to convert a long running operation into an RxJava Observable using
- How to convert a simulated "read database" operation into an
- How to create an
defer()lets you wrap an expensive method call
- The expensive method call will only happen when something subscribes
- Example project reading a value from a database, done with the use of
defer(), to do the read in a background thread.
- How to handle exceptions when using
Other Relevant Links:
Will you please explain why in spite of using SubscribeOn(Schedulers.io()) that heavy operation is not working on the io thread?
Hi Farhan, good question.
I assume your question is about the code at the ~3m 20s mark, where it's showing
The reason that doesn't work is, despite having
subscribeOn(Schedulers.io()) afterwards, the
Observable.just(Database.readValue()) command is executed immediately in the current thread; the current thread being the main thread in this instance.
This code doesn't work as the heavy operation actually happens during the creation of the
Observable. Once you have the
Observable created, you'll be able to switch schedulers, but the creation of the
Observable will happen immediately on the current thread (which is the main thread in the example).