Swift Get Next Page from header of NSHTTPURLResponse - swift

Question

I am consuming an API that gives me the next page in the Header inside a field called Link. (For example Github does the same, so it isn't weird.Github Doc)
The service that I am consuming retrieve me the pagination data in the following way:
As we can see in the "Link" gives me the next page,
With $0.response?.allHeaderFields["Link"]: I get </api/games?page=1&size=20>; rel="next",</api/games?page=25&size=20>; rel="last",</api/games?page=0&size=20>; rel="first".
I have found the following code to read the page, but it is very dirty... And I would like if anyone has dealt with the same problem or if there is a standard way of face with it. (I have also searched if alamofire supports any kind of feature for this but I haven't found it)
// MARK: - Pagination
private func getNextPageFromHeaders(response: NSHTTPURLResponse?) -> String? {
if let linkHeader = response?.allHeaderFields["Link"] as? String {
/* looks like:
<https://api.github.com/user/20267/gists?page=2>; rel="next", <https://api.github.com/user/20267/gists?page=6>; rel="last"
*/
// so split on "," the on ";"
let components = linkHeader.characters.split {$0 == ","}.map { String($0) }
// now we have 2 lines like '<https://api.github.com/user/20267/gists?page=2>; rel="next"'
// So let's get the URL out of there:
for item in components {
// see if it's "next"
let rangeOfNext = item.rangeOfString("rel=\"next\"", options: [])
if rangeOfNext != nil {
let rangeOfPaddedURL = item.rangeOfString("<(.*)>;", options: .RegularExpressionSearch)
if let range = rangeOfPaddedURL {
let nextURL = item.substringWithRange(range)
// strip off the < and >;
let startIndex = nextURL.startIndex.advancedBy(1) //advance as much as you like
let endIndex = nextURL.endIndex.advancedBy(-2)
let urlRange = startIndex..<endIndex
return nextURL.substringWithRange(urlRange)
}
}
}
}
return nil
}

Answer 1

I think that the forEach() could have a better solution, but here is what I got:
let linkHeader = "</api/games?page=1&size=20>; rel=\"next\",</api/games?page=25&size=20>; rel=\"last\",</api/games?page=0&size=20>; rel=\"first\""
let links = linkHeader.components(separatedBy: ",")
var dictionary: [String: String] = [:]
links.forEach({
let components = $0.components(separatedBy:"; ")
let cleanPath = components[0].trimmingCharacters(in: CharacterSet(charactersIn: "<>"))
dictionary[components[1]] = cleanPath
})
if let nextPagePath = dictionary["rel=\"next\""] {
print("nextPagePath: \(nextPagePath)")
}
//Bonus
if let lastPagePath = dictionary["rel=\"last\""] {
print("lastPagePath: \(lastPagePath)")
}
if let firstPagePath = dictionary["rel=\"first\""] {
print("firstPagePath: \(firstPagePath)")
}
Console output:
$> nextPagePath: /api/games?page=1&size=20
$> lastPagePath: /api/games?page=25&size=20
$> firstPagePath: /api/games?page=0&size=20
I used components(separatedBy:) instead of split() to avoid the String() conversion at the end.
I created a Dictionary for the values to hold and removed the < and > with a trim.