展平Scala中嵌套的Json对象列表(Flattening a List of nested Json objects in Scala)

 首先让我们解决分页问题。 现在分页不可撤销，但它确实要求您的服务器接受页面，以便它知道返回哪组结果。 例如，如果您的组大小为20，则有100个结果，在您的网址请求中，您必须指定是否需要第0 - 4页。 
 
 我个人建议使用一个开源库来为你处理分页。 看看MMRecord ，它确实为你提供了一些分页支持。 如果您有野心想要并且想要自己实现它，那么您需要构建一些网络服务，最好建立在AFNetworking之类的内容上 ，以跟踪最后请求的页面。 
 
 在swift中，这看起来像这样： 
 
 NetworkManager.GET('http://someURI', parameters: [:], page: 0) 
 
 不知何故，你必须保持下一个页面的状态。 如此处所建议的，更好的设计是使用协议和某种数据源包装器，它从客户端抽象出机制。 
 
 现在，您还询问了如何检测表视图何时滚动到末尾，以便您可以触发网络请求。 一种方法是简单观察scrollViewDidScroll （ UITableView继承自UIScrollView ）并检查contentOffset是否>= self.tableView.contentSize - self.tableView.frame.size.height （这将是当时的内容偏移量）表一直滚动）。 但是，您希望应用程序的用户能够继续滚动而不必滚动到底部，等待它加载，然后继续滚动，这样当我们向页面向下滚动3/4时我会发出请求如果您知道请求将需要很长时间，甚至中途。 这样，当它们到达底部时，它已经加载了数据，用户永远不会知道差异。 
 
 编辑： 
 
 除了我的回答，我会为你添加一些建议。 1）看看这个： http ： //artsy.github.io/blog/2015/09/24/mvvm-in-swift/ 
 
 它讨论了MVVM设计模式，它更适合保持状态而不是视图控制器中的状态变量。 
 
 2）你的函数， fetchAllObjectsFromParse是一个坏名字，原因有两个。 首先，该名称暗示该函数一次性获取所有对象，如果您希望对其进行分页，则不是这种情况。 
 
 也许是这样的： 
 
func fetchObjectsForModel(model: String, atPage page: Int, withBlock block: ([PFObejct], NSError?) -> Void {

}
 
 然后创建另一个将生成您的查询的函数： 
 
func generateQueryWithSkips(skip: Int, limit: Int) -> PFQuery {}

First let's tackle pagination. Now pagination isn't undoable but it does require your server to accept pages so that it knows which set of results to return. For example if you have 100 results in group sizes of twenty, in your url request you'd have to specify if you wanted page 0 - 4. 
 
I would personally recommend using an open-source library to take care of pagination for you. Take a look at MMRecord, it does have some pagination support for you. If you are feeling ambitious and you want to implement it on your own, you'll need to build some networking service, best built on something like AFNetworking that keeps track of the last requested page.
 
In swift, that would look something like this:
 
NetworkManager.GET('http://someURI', parameters: [:], page: 0)
 
Somehow you have to keep the state of which page is next around. A better design, as suggested here, is to use protocols and some sort of datasource wrapper that abstracts the mechanism from the client.
 
Now you also asked how to detect when a table view scrolls to the end so you can fire off a network request. One way to do it is simple observe scrollViewDidScroll (UITableView inherits from UIScrollView) and check to see if the contentOffset is >= self.tableView.contentSize - self.tableView.frame.size.height (which will be the content offset of when the table is all the way scrolled). However you want the user of your application to be able to keep scrolling without having to scroll to the bottom, waiting for it to load, then continue scrolling so what I would is fire the request when they've scrolled 3/4 down the page or even halfway if you know the request will take a long time. That way by the time they get to the bottom, it'll already have loaded the data and the user will never know the difference. 
 
EDIT:
 
I would add a few suggestions for you apart from my answer. 1) Take a look at this: http://artsy.github.io/blog/2015/09/24/mvvm-in-swift/
 
It talks about the MVVM design pattern which is much better for holding state instead of having state variables in your view controller. 
 
2) your function, fetchAllObjectsFromParse is a bad name for a couple of reasons. First, the name implies that the function fetches all the objects at once, which isn't the case if you want it paginated.
 
Maybe something like this:
 
func fetchObjectsForModel(model: String, atPage page: Int, withBlock block: ([PFObejct], NSError?) -> Void {

}
 
And then make another function which will generate your query:
 
func generateQueryWithSkips(skip: Int, limit: Int) -> PFQuery {}