Android Jetpack: LiveData
Articles Blog

Android Jetpack: LiveData

August 11, 2019


LYLA FUJIWARA: At
Google I/O 2017, we announced the
Lifecycle library, a set of libraries and guidance
for avoiding memory leaks and solving common Android
lifecycle challenges. The Lifecycle
library has hit 2.0. It’s also now part of Jetpack
and includes new integrations with Databinding. This is a tour of the Lifecycle
library’s LiveData class. LiveData is an observable
data holder class that is also lifecycle aware. Let’s take a look at an example. You’re going to have
your UI, and then you’re also going to have this
LiveData object, which holds some data that you
want to show on screen. The UI observes the
LiveData object. This is like saying the UI
wants to be notified of updates. Therefore, when the
LiveData changes, the UI will get
notified, and then the UI can redraw itself
with the new data. So in short, LiveData
makes it easy to keep what’s going on screen
in sync with the data. OK, so here’s some actual code. LiveData objects will usually
be kept in the ViewModel class. If you’re not sure what a
ViewModel is, no worries. Check out the video. Let’s say you’re creating
an activity and ViewModel for a user profile. You’ll have this
user LiveData object that holds a User object. Now, over in your
activity’s onCreate, you’ll get that LiveData
from the ViewModel class. Call observe on the LiveData. For the first argument, you’re
going to pass in the UI, in this case the activity. The second argument
is an “observer,” which is just a callback. Here you will call the
code to update the UI. Now you can change the
LiveData by calling either setValue or postValue. Use setValue if you’re
running on the UI thread, and then use postValue if you’re
running on a background thread. When either setValue
or postValue is called, the LiveData notifies
active observers. If you’re using Android
Studio 3.1 or higher, LiveData and ViewModels
work with Databinding. Usually you’re going to go
ahead and bind your View Model to your XML layout, which is
explained in the ViewModel video. Now, after associating your
ViewModel and Databinding layout, you just need to
add this single line change to have your LiveData
be properly observed when bound to the XML. You can now include references
in your XML to your ViewModel and the LiveData stored with it. If you use Databinding,
you’re going to no longer need to
actually manually set up these observers. So instead of creating
this LiveData observer code that I showed you
before, you could remove all that boilerplate. Instead, the TextView’s
XML references the LiveData directly. What makes LiveData different
from other observables is that it is also
lifecycle aware. This means that it understands
whether your UI is on screen, offscreen or destroyed. LiveData knows
about your UI state because you passed it
when you call Observe. OK, so here’s some benefits
of LiveData’s life cycle awareness. So let’s say your
activity is not on screen, then your LiveData doesn’t
trigger useless UI updates. If the activity or
UI gets destroyed, then the observation connection
is cleaned up automatically by LiveData. Thus you’ll never
accidentally trigger an activity or fragment that
is offscreen or destroyed to redraw itself. This is possible in part because
of interfaces and classes in the Lifecycle
library that are also used by framework classes. These classes are– Lifecycle, which is an object
that represents an Android lifecycle and what
state it’s in; LifecycleOwner, which is an
interface for objects that have a lifecycle like
AppCompatActivity or an activity fragment; and
finally, LifecycleObserver, which is an interface for
observing a lifecycle. OK, so LiveData is a
lifecycle observer. It abstracts away
the need for you to deal directly with activity
or factored lifecycle. So those are the basics
of working with LiveData and why it’s useful. I’m going to touch on a
few more complex usages. Room is built to work
well with LiveData. Room can return LiveData
objects which are automatically notified when the database data
changes and have their data loaded in a background thread. This makes it easy to
have the UI update when your database updates. You can go ahead and learn more
about Room in the introduction to Room video. LiveData also provides
transformations, including map, switchMap and a class
called MediatorLiveData for your own custom
transformations. Map lets you apply a function
to the output of LiveData A and then propagate the results
downstream to LiveData B. For example, you could use
LiveData to take a user object and instead output a string
of the user’s combined first and last name. SwitchMap function
transformation is a lot like map, but
for mapping functions that emit LiveData
instead of values. So an example here is if
you have a bunch of users, perhaps stored in
a Room database, you might have a lookup
function for those users. Using switchMap, you’d have
a LiveData for the user ID. Whenever the ID changes,
your user lookup function would be called with that ID. The result LiveData now
references the newly found user LiveData. OK, so no matter how
many different times you call this look up function
and get a different LiveData, your UI only needs to observe
the result LiveData once, which is the power of switchMap. Now, if you want to go ahead
and make your own custom data transformations, you
should take a look at the MediatorLiveData class. MediatorLiveData includes
methods to add and remove source LiveData objects. You could then combine
and propagate events from all these
sources downstream. Getting started with
LiveData is simple, but there is a lot of
potential for experimentation with this lifecycle
aware observable. Hopefully this has inspired
you to build something. And as always, the
documentation is linked.

Only registered users can comment.

  1. While Kotlin may be a good functional object oriented programming language, a vast majority of android developers are Java developers and among them, a lot of them use Java for other purposes as well. This means that all these people(at least the vast majority) are more familiar with the imperative style of programming rather than functional.
    It would be rather beneficial for the majority if the sample code in the videos are in Java. A change this fast is quiet difficult to adapt to.
    Thank You.

  2. All these great new features coming but examples on kotlin, becoming really hard to relate to this. been developing android apps for years but still, seeing the unfamiliar code of kotlin makes it kinda harder to instantly pick the concept and grasp it. Along with learning the concept i feel like i have to learn a whole language as well, making things more complicated for no reason.

  3. Kotlin's been "officially supported language for Android development" since 2017's I/O, it's been a year, surely we can pick it up by now 😀

  4. it took me 3 minutes to figure out that the person speaking is a women 😀 i am slow as Android Studio in my machine

  5. Kotlin > Java. Period. Its way less cluttered than Java. I have been programming Android apps for 3 months using Kotlin. I am really happy with it.

  6. In 00:40 instead of saying the UI will be notified, you should have said the observer method will get notified.
    Maybe the programmer wants to do something else instead of changing the UI.
    What do you think about it?

    I appreciate your suggestions.

  7. If you're going to deprecate things like loaders, either use Java in your slides or use pseudocode. You told me in a video I should be using loaders in Java, so why isn't your retraction in Java too? As my ASD is likely to prevent me from ever understanding Kotlin as it lacks the syntax of other programming languages, YouTube search suggests my only option to understand LiveData is to learn Hindi.

    Perhaps a random video in Hindi will do a better job at explaining it than you are, because unlike spoken language, programming languages like C, C#, Java, Python, and PHP, are all somewhat similar visually, and while YouTube may be a tad awful at subtitling videos, there's more chance of Auto-Translate being intelligible than Google releasing the Java slides.

    If you're going to insist on using Kotlin in all your slides, please add programming languages to Google Translate's live image translation feature. At the moment you are like a country with two official languages that is discriminating against people that only understand one of the languages… so a bit like Cameroon.

    Linguicism: "ideologies and structures which are used to legitimate, effectuate, and reproduce unequal division of power and resources (both material and non-material) between groups which are defined on the basis of language" –Tove Skutnabb-Kangas

    "There are different forms of linguicism. Overt linguicism is exemplified by the prohibition of the use of particular languages for instruction. Covert linguicism is illustrated by de facto non use of certain languages as languages of instruction, even if their use is not explicitly forbidden." –William Velez

  8. kindly do more effort to explain the new introduced concepts in both java and kotlin .. and don't act like you don't care about java developers anymore .. this is disrespectful. both have the right to get detailed explanations in both your live shows and learning videos.. not everyone have the luxury to learn a new programming language just to integrate new introduced components to an existing project .. and if your documentation do so .. it is just Not enough as we deal with more serious stuff in the real world.. the docs only shows basic stuff.
    #google #javadevelopers #androidjetpack #androidarchitecturecomponents
    Regards

  9. A lot of good things about live data, but it's actually very bad.

    – No normal error handling

    — If 'observe' is not used in the 'onCreate' method, then there will be problems when change screen orientation

Leave a Reply

Your email address will not be published. Required fields are marked *