Reputation: 409
I am trying to find a way to update single item in recycler view using PagingAdapter from Paging 3 library. I have found only one way with PagingAdapter.refresh() method. But this method force to load all list from network. Does anybody know how to implement it without loading all pages from network?
Upvotes: 23
Views: 16851
Reputation: 1332
To accomplish this (note: here i don't use database for cache, only remote), we can create a mutable Flow of data list in the ViewModel that contains the items we want to manipulate. When observing the Flow of Paging 3, we can combine it with our local Flow and find the item we want to change. We can then change it before submitting it to the view to observe.
Here is an example of a simplified ViewModel that demonstrates this solution:
class ExampleViewModel : ViewModel() {
private val _localDataList = MutableStateFlow(listOf<MyData>())
// Observe the Paging 3 Flow and combine it with the local Flow
val combinedDataList = paging3Flow.cachedIn(viewModelScope).combine(_localDataList) { paging, local ->
// Find and update the desired item in the list
paging.map {
if (it.id == local.id) local
else it
}
}
// Method to update the item in the local Flow
fun updateItem(item: MyData) {
val updatedItem = getUpdatedItemFromServerUseCase()
val newList = _localDataList.value.filterNot { it.id == updatedItem.id } // remove old version if any.
_localDataList.value = newList + updatedItem
}
}
If you want to remove items from the list, you can create a new data class that contains the item with an updateType enum that contains update and remove values. Depending on the update type, you can map the PagingData accordingly.
Here is an example of how you can use the updateType enum:
enum class UpdateType {
UPDATE, REMOVE
}
data class UpdateData(
val updateType: UpdateType,
val item: MyData
)
class ExampleViewModel : ViewModel() {
private val _localDataList = MutableStateFlow(listOf<UpdateData>())
// Observe the Paging 3 Flow and combine it with the local Flow
val combinedDataList = paging3Flow.combine(_localDataList.) { paging, local ->
// Find and update or remove the desired item in the list
paging.map {
local.find { localItem -> localItem.item.id == it.id }?.let { localItem ->
when (localItem.updateType) {
UpdateType.UPDATE -> localItem.item
UpdateType.REMOVE -> null
}
} ?: it
}.filter { it!=null }
}
// Method to remove the item from the local Flow
fun updateItem(item: MyData) {
removeItemFromServerUseCase()
val newList = _localDataList.value.filterNot { it.item.id == item.id } // remove old version if any.
_localDataList.value = newList + UpdateData(UpdateType.REMOVE, updatedItem)
}
}
Upvotes: 4
Reputation: 93
(adapterComment.snapshot().items as MutableList<Model>)[position].likeStatus = 0
adapterComment.notifyItemChanged(position)
Upvotes: -1
Reputation: 1834
Example :
fun markItemAsRead(position: Int) {
snapshot()[position].read = true
notifyItemChanged(position)
}
Upvotes: 2
Reputation: 3905
Currently, the only way to update the backing dataset is to invalidate and reload the list. This is generally an acceptably cheap option for layered sources that use a cached layer (either in db such as room or in memory), although there is ongoing work to support more granular updates (see https://issuetracker.google.com/160232968).
In terms of layered source for now, you'll need to move your network calls into a RemoteMediator
which you can register in Pager
's constructor, and cache your network fetches into either a DB like with Room (which can generate a PagingSource
implementation for you), or write an in-memory one yourself.
The codelab and DAC docs are a great resource for this, and have code samples to guide you!
Upvotes: 22