I have a lot of places in the code where Alamofire request/response are handled.
Each of this requests may fail because of some intermittent problem (the most common is flaky network).
I would like to be able to retry requests 3 times before bailing out.
The straightforward method would be to having something like that
var errorCount = 0
func requestType1() {
let request = Alamofire.request(...).responseJSON { response in
if (isError(response) && errorCount < 3) {
errorCount += 1
request1()
}
if (isError(response)) {
handleError()
}
handleSuccess()
}
}
However, I dislike this approach A LOT for multiple reasons. The most obvious is that I will need to implement such code for each request type (and I have something like 15 of them).
I am curios whether there is way to do something like (where the changes are minimal and non intrusive)
let request = Alamofire.request(..., **3**)
One of the bits of syntactic sugar you get with Swift is you can use this:
public func updateEvents(someNormalParam: Bool = true, someBlock: (Void->Void))
Like this:
updateEvents(someNormalParam: false) {...}
Note the block is outside the () of the updateEvents function, contrary to where you'd normally expect it. It works only if the block is the last thing in the declaration of the function.
That means if you happen to have a block such as your Alamofire request, you can effectively wrap it with your retry functionality. One slightly complicating issue is you want to call a block within the block. Not a big deal:
func retryWrapper(alamoBlock: (Void->Request)) {
alamoblock().responseJSON() {
//Your retry logic here
}
}
And you use it like so:
retryWrapper() {
Alamofire.request(method, targetUrl, parameters: parameters, encoding: encoding)
}
Meaning all you have to do is find your Alamofire calls and wrap them in { } and put retryWrapper() before. The retry logic itself is only there once.