[英]How to add date separators in recycler view using Paging Library?
After a lot of searching, I know its possible with regular adapter, but I have no idea how to do it using Paging Library.经过大量搜索,我知道使用常规适配器是可能的,但我不知道如何使用分页库来做到这一点。 I don`t need code just a clue.我不需要代码只是一个线索。
Example例子
To add separators, you essentially have 2 options:要添加分隔符,您基本上有两个选项:
For the paging library only option 2 is viable since it only partially loads the data and inserting the separators becomes much more complicated.对于分页库,只有选项 2 是可行的,因为它仅部分加载数据并且插入分隔符变得更加复杂。 You will simply need to figure out a way to check if item x
is a different day than item x-1
and show/hide the date section in the view depending on the result.您只需要找出一种方法来检查项目x
是否与项目x-1
并根据结果在视图中显示/隐藏日期部分。
I was in the same spot as you and I came up with this solution.我和你在同一个地方,我想出了这个解决方案。
One important note though, in order to implement this I had to change my date converter to the database, from long to string to store a timestamp一个重要的注意事项,为了实现这一点,我不得不将我的日期转换器更改为数据库,从 long 到 string 以存储时间戳
these are my converters这些是我的转换器
class DateConverter {
companion object {
@JvmStatic
val formatter = SimpleDateFormat("yyyyMMddHHmmss", Locale.ENGLISH)
@TypeConverter
@JvmStatic
fun toDate(text: String): Date = formatter.parse(text)
@TypeConverter
@JvmStatic
fun toText(date: Date): String = formatter.format(date)
}
}
Some starting info though, I have a list of report headers that I wish to show , and page through and be able to filter虽然有一些起始信息,但我有一个我希望显示的报告标题列表,可以翻页并能够过滤
They are represented by this object:它们由这个对象表示:
data class ReportHeaderEntity(
@ColumnInfo(name = "id") override val id: UUID
, @ColumnInfo(name = "name") override val name: String
, @ColumnInfo(name = "description") override val description: String
, @ColumnInfo(name = "created") override val date: Date)
I also wanted to add separators between the items in the list to show them by date我还想在列表中的项目之间添加分隔符以按日期显示它们
I achieved this by doing the following:我通过执行以下操作实现了这一点:
I created a new query in room like this我在这样的房间中创建了一个新查询
@Query(
"SELECT id, name, description,created " +
"FROM (SELECT id, name, description, created, created AS sort " +
" FROM reports " +
" WHERE :filter = '' " +
" OR name LIKE '%' || :filter || '%' " +
" OR description LIKE '%' || :filter || '%' " +
" UNION " +
" SELECT '00000000-0000-0000-0000-000000000000' as id, Substr(created, 0, 9) as name, '' as description, Substr(created, 0, 9) || '000000' AS created, Substr(created, 0, 9) || '256060' AS sort " +
" FROM reports " +
" WHERE :filter = '' " +
" OR name LIKE '%' || :filter || '%' " +
" OR description LIKE '%' || :filter || '%' " +
" GROUP BY Substr(created, 0, 9)) " +
"ORDER BY sort DESC ")
fun loadReportHeaders(filter: String = ""): DataSource.Factory<Int, ReportHeaderEntity>
This basically creates a separator line for all the items I have filtered through这基本上为我过滤的所有项目创建了一个分隔线
it also creates a dummy date for sorting (with the time of 25:60:60 so that it will always appear in front of the other reports)它还创建了一个用于排序的虚拟日期(时间为 25:60:60,以便它始终出现在其他报告的前面)
I then combine this with my list using union and sort them by the dummy date然后我使用 union 将它与我的列表结合起来,并按虚拟日期对它们进行排序
The reason I had to change from long to string is because it is much easier to create dummy dates with string in sql and seperate the date part from the whole date time我必须从 long 更改为 string 的原因是因为在 sql 中使用 string 创建虚拟日期并将日期部分与整个日期时间分开要容易得多
The above creates a list like this:上面创建了一个这样的列表:
00000000-0000-0000-0000-000000000000 20190522 20190522000000
e3b8fbe5-b8ce-4353-b85d-8a1160f51bac name 16769 description 93396 20190522141926
6779fbea-f840-4859-a9a1-b34b7e6520be name 86082 description 21138 20190522141925
00000000-0000-0000-0000-000000000000 20190521 20190521000000
6efa201f-d618-4819-bae1-5a0e907ddcfb name 9702 description 84139 20190521103247
In my PagedListAdapter I changed it to be an implementation of PagedListAdapter<ReportHeader, RecyclerView.ViewHolder>
(not a specific viewholder)在我的 PagedListAdapter 中,我将其更改为PagedListAdapter<ReportHeader, RecyclerView.ViewHolder>
(不是特定的viewholder)的实现
Added to the companion object:添加到伴生对象:
companion object {
private val EMPTY_ID = UUID(0L,0L)
private const val LABEL = 0
private const val HEADER = 1
}
and overrode get view type like so:并像这样覆盖 get 视图类型:
override fun getItemViewType(position: Int): Int = if (getItem(position)?.id ?: EMPTY_ID == EMPTY_ID) LABEL else HEADER
I then created two seperate view holders :然后我创建了两个单独的视图持有者:
class ReportHeaderViewHolder(val binding: ListItemReportBinding) : RecyclerView.ViewHolder(binding.root)
class ReportLabelViewHolder(val binding: ListItemReportLabelBinding) : RecyclerView.ViewHolder(binding.root)
and implemented the other overriden methods like so:并实现了其他覆盖方法,如下所示:
override fun onCreateViewHolder(parent: ViewGroup, viewType: Int): RecyclerView.ViewHolder {
val inflater = LayoutInflater.from(parent.context)
return when (viewType) {
HEADER -> ReportHeaderViewHolder(DataBindingUtil.inflate(inflater, R.layout.list_item_report, parent, false))
else -> ReportLabelViewHolder(DataBindingUtil.inflate(inflater, R.layout.list_item_report_label, parent, false))
}
}
override fun onBindViewHolder(holder: RecyclerView.ViewHolder, position: Int) {
val reportItem = getItem(position)
when (getItemViewType(position)) {
HEADER -> {
(holder as ReportHeaderViewHolder).binding.apply {
report = reportItem
executePendingBindings()
}
}
LABEL -> {
(holder as ReportLabelViewHolder).binding.apply {
date = reportItem?.name
executePendingBindings()
}
}
}
}
I hope this helps and inspires people to find even better solutions我希望这有助于并激励人们找到更好的解决方案
You can achieve the same result using insertSeparators
in Paging 3 library.您可以使用Paging 3库中的insertSeparators
获得相同的结果。 Make sure your items are sorted
by date.确保您的物品sorted
日期sorted
。
Inside or viewmodel
retrieve a Pager
something like that在内部或viewmodel
检索类似这样的Pager
private val communicationResult: Flow<PagingData<CommunicationHistoryItem>> = Pager(
PagingConfig(
pageSize = 50,
enablePlaceholders = false,
maxSize = 400,
initialLoadSize = 50
)
) {
CommunicationPagingSource(repository)
}.flow.cachedIn(viewModelScope)
After all insert separators
like a header毕竟像标题一样insert separators
val groupedCommunicationResult = communicationResult
.map { pagingData -> pagingData.map { CommunicationHistoryModel.Body(it) } }
.map {
it.insertSeparators{ after, before ->
if (before == null) {
//the end of the list
return@insertSeparators null
}
val afterDateStr = after?.createdDate
val beforeDateStr = before.createdDate
if (afterDateStr == null || beforeDateStr == null)
return@insertSeparators null
val afterDate = DateUtil.parseAsCalendar(afterDateStr)?.cleanTime()?.time ?: 0
val beforeDate = DateUtil.parseAsCalendar(beforeDateStr)?.cleanTime()?.time ?: 0
if (afterDate > beforeDate) {
CommunicationHistoryModel.Header( DateUtil.format(Date(beforeDate))) // dd.MM.yyyy
} else {
// no separator
null
}
}
}
cleanTime
is required for grouping
by dd.MM.yyyy
ignoring time cleanTime
需要grouping
通过dd.MM.yyyy
忽略时间
fun Calendar.cleanTime(): Date {
set(Calendar.HOUR_OF_DAY, 0)
set(Calendar.MINUTE, 0)
set(Calendar.SECOND, 0)
set(Calendar.MILLISECOND, 0)
return this.time
}
When binding the data pass in the previous item as well绑定数据时也传入前一项
override fun onBindViewHolder(holder: ViewHolder, position: Int) {
val item = getItem(position)
val previousItem = if (position == 0) null else getItem(position - 1)
holder.bind(item, previousItem)
}
Every view then sets a header, which is only made visible if the previous item doesn't have the same header.然后每个视图设置一个标题,只有在前一项没有相同标题时才可见。
val previousHeader = previousItem?.name?.capitalize().first()
val header = item?.name?.capitalize()?.first()
view.cachedContactHeader.text = header
view.cachedContactHeader.isVisible = previousHeader != header
As mentioned here , Paging Library works with PagedListAdapter . 如前所述这里 ,寻呼图书馆工作与PagedListAdapter 。 And PagedListAdapter extends from RecyclerView.Adapter, so you can do this simply like in Recyclerview ( example ). PagedListAdapter从RecyclerView.Adapter扩展,因此您可以像在Recyclerview( 示例 )中那样完成此操作。 Just in two words - you need to use different view types for your date headers and content items. 只用两个字 - 您需要为日期标题和内容项使用不同的视图类型。
Kiskae's answer is excellent and for your case option 2 probably works well. Kiskae 的回答非常好,对于您的情况,选项 2 可能效果很好。
In my case I wanted to have one additional item that wasn't in the database, like this:在我的情况下,我想要一个不在数据库中的附加项目,如下所示:
It needed to be clickable as well.它也需要可点击。 There's the usual way of overriding getItemCount
to return +1 and offsetting positions for the other methods.通常的方法是覆盖getItemCount
以返回 +1 并偏移其他方法的位置。
But I stumbled on another way that I haven't seen documented yet that might be useful for some cases.但是我偶然发现了另一种我还没有看到记录的方法,它可能对某些情况有用。 You might be able to incorporate additional elements into your query using union
:您可以使用union
将其他元素合并到您的查询中:
@Query("select '' as name, 0 as id " +
"union " +
"select name, id from user " +
"order by 1 asc")
DataSource.Factory<Integer, User> getAllDataSource();
That means the data source actually returns another item in the beginning, and there's no need to adjust positions.这意味着数据源实际上在开始时返回了另一个项目,无需调整位置。 In your adapter, you can check for that item and handle it differently.在您的适配器中,您可以检查该项目并以不同方式处理它。
In your case the query would have to be different but I think it would be possible.在您的情况下,查询必须有所不同,但我认为这是可能的。
声明:本站的技术帖子网页,遵循CC BY-SA 4.0协议,如果您需要转载,请注明本站网址或者原文地址。任何问题请咨询:yoyou2525@163.com.