i can print the pagedArray object content but when i do something like this pagedArray[0], the result is always nil and after debugging i get this : Array representation: [nil]

I have recently encountered several crashes related to this function

/// Accesses and sets elements for a given flat index position.
    /// Currently, setter can only be used to replace non-optional values.
    public subscript (position: Index) -> T? {
        get {
            let pageIndex = page(for: position)
            
            if let page = elements[pageIndex] {
                return page[position%pageSize]
            } else {
                // Return nil for all pages that haven't been set yet
                return nil
            }
        }
        
        set(newValue) {
            guard let newValue = newValue else { return }
            
            let pageIndex = page(for: position)
            var elementPage = elements[pageIndex]
            elementPage?[position % pageSize] = newValue
            elements[pageIndex] = elementPage
        }
    }

specifically this line return page[position%pageSize] throws IndexOutOfBounds errors

Version used: 0.8

Hi,
This looks like a great idea but I'm having some difficulty hocking the code to a real web service

Could please publish a working example.

In order to migrate and test our app to Swift 2.3/3.0 prior to the release of Xcode 8 and iOS 10, all of our dependencies need to be migrated first.

This can be solved by pointing our podfile to those respective Swift 2.3/3.0 branches, but this repo does not appear to have either of those branches.

This blog post contains helpful tips for migrating open source libraries.

How do we deal with memory warning issues when memory runs out because of setting so many new elements into the array?

Hello,

This looks great on paper. Unfortunately I couldn't try it because its not Swift 2.0 compatible.

Any chance we can see an update soon?

Thanks in advance and great work
Ace

I am trying to implement something similar with the Example.
I am setting the total count from the metadata of the first api call to get the first page.
however even though I am not scrolling yet, there are 5-6 api calls to get the first page.
Here is my code :

    fileprivate func loadDataIfNeededForRow(_ row: Int, withSearchText: String?) {
        var type: VehicleSearchType = .full
        if let _ = withSearchText {
            type = .filtered
        }
        let currentPage = pagedArray.page(for: row)
        if needsLoadDataForPage(currentPage, type: type) {
            loadDataForPage(currentPage, withSearchText: withSearchText)
        }
        
        let preloadIndex = row+PreloadMargin
        if preloadIndex < pagedArray.endIndex && shouldPreload {
            let preloadPage = pagedArray.page(for: preloadIndex)
            if preloadPage > currentPage && needsLoadDataForPage(preloadPage, type: type) {
                loadDataForPage(preloadPage, withSearchText: withSearchText)
            }
        }
    }
    
    private func needsLoadDataForPage(_ page: Int, type: VehicleSearchType) -> Bool {
        
        if type == .filtered {
            return filteredVehicleList.elements[page] == nil && dataLoadingOperations[page] == nil
        }
        return pagedArray.elements[page] == nil && dataLoadingOperations[page] == nil
    }

Hello

Just wondering if this library plays well with JSON data Object. If so, how'd you go about linking them together?

if it?
pagedArray.setElements(JSONobject.arrayobject, pageIndex: 1)

Hi,

I think you forgot to check the shared scheme chekbox in order to be Carthage compatible.

   func testPagedArraySubscriptAfterCountChanged() {
        var x = PagedArray<String>(count: 6, pageSize: 4, startPage: 1)
        x.set(["a", "b"], forPage: 2)
        x.count = 8
        let y = x[7] // crash
        XCTAssert(y == nil)
   }

It happens sometimes and I've found that best solution for me would be to make PagedArray bulletproof for that case. Could we have some simple solution like Here ? Or do you have some other suggestion ?

Hi,

Could you provide some example how to deal with API data is not providing count of all data?
For now it looks that everything is based on known count.

In comparison to using Realm. What are some tradeoffs and benefits of using this paging system?

https://realm.io/

Please can you provide simple example with alamofire, i am struggling with it

This is the test:

func testLoadedElementsForOnePageAndAHalfOfTheLastOne() {
var tinyArray: PagedArray = PagedArray(count: 15, pageSize: 10)
tinyArray.setElements(Array(0...9), page: 0)
tinyArray.setElements(Array(10...11), page: 1)
XCTAssertEqual(tinyArray.loadedElements.count, 12, "Array total count is not correct")
}

How can I handle the changes in the dataset ? I think this should work only if the data is not ordered or i'm i wrong ? If an ID is missing and added while you scroll you will have a shift in the rows

Hi, I am curious about the proper of connecting this library with the responses of my API. To be more clear, Our API sends back the data page by page which means I have to make a HTTP request for each page. I am looking for a library that could ease my job here, so do you think that I can use this repository? If yes, how? (the efficient way). Thanks