android

Pagination (Paging) is a common and powerful technique in Android Development when making HTTP requests or fetching data from the database. Eventyay Attendee has found many situations where data binding comes in as a great solution for our network calls with Retrofit. Let’s take a look at this technique.

• Problems without Pagination in Android Development
• Implementing Pagination with Kotlin with Retrofit
• Results and GIF
• Conclusions

PROBLEMS WITHOUT DATABINDING IN ANDROID DEVELOPMENT

Making HTTP requests to fetch data from the API is a basic work in any kind of application. With the mobile application, network data usage management is an important factor that affects the loading performance of the app. Without paging, all of the data are fetched even though most of them are not displayed on the screen. Pagination is a technique to load all the data in pages of limited items, which is much more efficient

IMPLEMENTING DATABINDING IN FRAGMENT VIEW

Step 1:  Set up dependency in build.gradle

// Paging
implementation "androidx.paging:paging-runtime:$paging_version"
implementation "androidx.paging:paging-rxjava2:$paging_version"

Step 2:  Set up retrofit to fetch events from the API

@GET("events?include=event-sub-topic,event-topic,event-type")
fun searchEventsPaged(
   @Query("sort") sort: String,
   @Query("filter") eventName: String,
   @Query("page[number]") page: Int,
   @Query("page[size]") pageSize: Int = 5
): Single<List<Event>>

Step 3: Set up the DataSource

DataSource is a base class for loading data in the paging library from Android. In Eventyay, we use PageKeyedDataSource. It will fetch the data based on the number of pages and items per page with our default parameters. With PageKeyedDataSource, three main functions loadInitial(), loadBefore(), loadAfter() are used to to load each chunks of data.

class EventsDataSource(
   private val eventService: EventService,
   private val compositeDisposable: CompositeDisposable,
   private val query: String?,
   private val mutableProgress: MutableLiveData<Boolean>

) : PageKeyedDataSource<Int, Event>() {
   override fun loadInitial(
       params: LoadInitialParams<Int>,
       callback: LoadInitialCallback<Int, Event>
   ) {
       createObservable(1, 2, callback, null)
   }

   override fun loadAfter(params: LoadParams<Int>, callback: LoadCallback<Int, Event>) {
       val page = params.key
       createObservable(page, page + 1, null, callback)
   }

   override fun loadBefore(params: LoadParams<Int>, callback: LoadCallback<Int, Event>) {
       val page = params.key
       createObservable(page, page - 1, null, callback)
   }

   private fun createObservable(
       requestedPage: Int,
       adjacentPage: Int,
       initialCallback: LoadInitialCallback<Int, Event>?,
       callback: LoadCallback<Int, Event>?
   ) {
       compositeDisposable +=
           eventService.getEventsByLocationPaged(query, requestedPage)
               .withDefaultSchedulers()
               .subscribe({ response ->
                   if (response.isEmpty()) mutableProgress.value = false
                   initialCallback?.onResult(response, null, adjacentPage)
                   callback?.onResult(response, adjacentPage)
               }, { error ->
                   Timber.e(error, "Fail on fetching page of events")
               }
           )
   }
}

Step 4: Set up the Data Source Factory

DataSourceFactory is the class responsible for creating DataSource object so that we can create PagedList (A type of List used for paging) for events.

class EventsDataSourceFactory(
   private val compositeDisposable: CompositeDisposable,
   private val eventService: EventService,
   private val query: String?,
   private val mutableProgress: MutableLiveData<Boolean>
) : DataSource.Factory<Int, Event>() {
   override fun create(): DataSource<Int, Event> {
       return EventsDataSource(eventService, compositeDisposable, query, mutableProgress)
   }
}

Step 5: Adapt the current change to the ViewModel. 

Previously, events fetched in List<Event> Object are now should be turned into PagedList<Event>.

sourceFactory = EventsDataSourceFactory(
   compositeDisposable,
   eventService,
   mutableSavedLocation.value,
   mutableProgress
)
val eventPagedList = RxPagedListBuilder(sourceFactory, config)
   .setFetchScheduler(Schedulers.io())
   .buildObservable()
   .cache()

compositeDisposable += eventPagedList
   .subscribeOn(Schedulers.io())
   .observeOn(AndroidSchedulers.mainThread())
   .distinctUntilChanged()
   .doOnSubscribe {
       mutableProgress.value = true
   }.subscribe({
       val currentPagedEvents = mutablePagedEvents.value
       if (currentPagedEvents == null) {
           mutablePagedEvents.value = it
       } else {
           currentPagedEvents.addAll(it)
           mutablePagedEvents.value = currentPagedEvents
       }
   }, {
       Timber.e(it, "Error fetching events")
       mutableMessage.value = resource.getString(R.string.error_fetching_events_message)
   })

Step 6: Turn ListAdapter into PagedListAdapter

PageListAdapter is basically the same ListAdapter to update the UI of the events item but specifically used for Pagination. In here, List objects can also be null.

class EventsListAdapter : PagedListAdapter<Event, EventViewHolder>(EventsDiffCallback()) {

   var onEventClick: EventClickListener? = null
   var onFavFabClick: FavoriteFabClickListener? = null
   var onHashtagClick: EventHashTagClickListener? = null

   override fun onCreateViewHolder(parent: ViewGroup, viewType: Int): EventViewHolder {
       val binding = ItemCardEventsBinding.inflate(LayoutInflater.from(parent.context), parent, false)
       return EventViewHolder(binding)
   }

   override fun onBindViewHolder(holder: EventViewHolder, position: Int) {
       val event = getItem(position)
       if (event != null)
           holder.apply {
               bind(event, position)
               eventClickListener = onEventClick
               favFabClickListener = onFavFabClick
               hashTagClickListAdapter = onHashtagClick
           }
   }

   /**
    * The function to call when the adapter has to be cleared of items
    */
   fun clear() {
       this.submitList(null)
   }

AND HERE ARE THE RESULTS…

CONCLUSION

Databinding is the way to go when working with a complex UI in Android Development. This helps reducing boilerplate code and to increase the readability of the code and the performance of the UI. One problem with data-binding is that sometimes, it is pretty hard to debug with unhelpful log messages. Hopefully, you can empower your UI in your project now with data-binding. 

Pagination is the way to go for fetching items from the API and making infinite scrolling. This helps reduce network usage and improve the performance of Android applications. And that’s it. I hope you can make your application more powerful with pagination. 

RESOURCES

Open Event Codebase: https://github.com/fossasia/open-event-attendee-android/pull/2012

Documentation: https://developer.android.com/topic/libraries/architecture/paging/ 

Google Codelab: https://codelabs.developers.google.com/codelabs/android-paging/#0