--- title: Fetching website data into memory framework: foundation role: article role_heading: Article path: foundation/fetching-website-data-into-memory --- # Fetching website data into memory Receive data directly into memory by creating a data task from a URL session. ## Overview Overview For small interactions with remote servers, you can use the URLSessionDataTask class to receive response data into memory (as opposed to using the URLSessionDownloadTask class, which stores the data directly to the file system). A data task is ideal for uses like calling a web service endpoint. You use a URL session instance to create the task. If your needs are fairly simple, you can use the shared instance of the URLSession class. If you want to interact with the transfer through delegate callbacks, you’ll need to create a session instead of using the shared instance. You use a URLSessionConfiguration instance when creating a session, also passing in a class that implements URLSessionDelegate or one of its subprotocols. Sessions can be reused to create multiple tasks, so for each unique configuration you need, create a session and store it as a property. note: Be careful to not create more sessions than you need. For example, if you have several parts of your app that need a similarly configured session, create one session and share it among them. Once you have a session, you create a data task with one of the dataTask() methods. Tasks are created in a suspended state, and can be started by calling resume(). Receive results with a completion handler The simplest way to fetch data is to create a data task that uses a completion handler. With this arrangement, the task delivers the server’s response, data, and possibly errors to a completion handler block that you provide. shows the relationship between a session and a task, and how results are delivered to the completion handler. To create a data task that uses a completion handler, call the dataTask(with:) method of URLSession. Your completion handler needs to do three things: Verify that the error parameter is nil. If not, a transport error has occurred; handle the error and exit. Check the response parameter to verify that the status code indicates success and that the MIME type is an expected value. If not, handle the server error and exit. Use the data instance as needed. The following example shows a startLoad() method for fetching a URL’s contents. It starts by using the URLSession class’s shared instance to create a data task that delivers its results to a completion handler. After checking for local and server errors, this handler converts the data to a string, and uses it to populate a WKWebView outlet. Of course, your app might have other uses for fetched data, like parsing it into a data model. Creating a completion handler to receive data-loading results func startLoad() { let url = URL(string: "https://www.example.com/")! let task = URLSession.shared.dataTask(with: url) { data, response, error in if let error = error { self.handleClientError(error) return } guard let httpResponse = response as? HTTPURLResponse, (200...299).contains(httpResponse.statusCode) else { self.handleServerError(response) return } if let mimeType = httpResponse.mimeType, mimeType == "text/html", let data = data, let string = String(data: data, encoding: .utf8) { DispatchQueue.main.async { self.webView.loadHTMLString(string, baseURL: url) } } } task.resume() } important: The completion handler is called on a different Grand Central Dispatch queue than the one that created the task. Therefore, any work that uses data or error to update the UI — like updating webView — should be explicitly placed on the main queue, as shown here. Receive transfer details and results with a delegate For a greater level of access to the task’s activity as it proceeds, when creating the data task, you can set a delegate on the session, rather than providing a completion handler. shows this arrangement. With this approach, portions of the data are provided to the urlSession(_:dataTask:didReceive:) method of URLSessionDataDelegate as they arrive, until the transfer finishes or fails with an error. The delegate also receives other kinds of events as the transfer proceeds. You need to create your own URLSession instance when using the delegate approach, rather than using the URLSession class’s simple shared instance. Creating a new session allows you to set your own class as the session’s delegate, as shown in the following example. Declare that your class implements one or more of the delegate protocols (URLSessionDelegate, URLSessionTaskDelegate, URLSessionDataDelegate, and URLSessionDownloadDelegate). Then create the URL session instance with the initializer init(configuration:delegate:delegateQueue:). You can customize the configuration instance used with this initializer. For example, it’s a good idea to set waitsForConnectivity to true. That way, the session waits for suitable connectivity, rather than failing immediately if the required connectivity is unavailable. Creating a URLSession that uses a delegate private lazy var session: URLSession = { let configuration = URLSessionConfiguration.default configuration.waitsForConnectivity = true return URLSession(configuration: configuration, delegate: self, delegateQueue: nil) }() The following example shows a startLoad() method that uses this session to start a data task, and uses delegate callbacks to handle received data and errors. This listing implements three delegate callbacks: urlSession(_:dataTask:didReceive:completionHandler:) verifies that the response has a succesful HTTP status code, and that the MIME type is text/html or text/plain. If either of these is not the case, the task is canceled; otherwise, it’s allowed to proceed. urlSession(_:dataTask:didReceive:) takes each Data instance received by the task and appends it to a buffer called receivedData. urlSession(_:task:didCompleteWithError:) first looks to see if a transport-level error has occurred. If there is no error, it attempts to convert the receivedData buffer to a string and set it as the contents of webView. Using a delegate with a URL session data task var receivedData: Data? func startLoad() { loadButton.isEnabled = false let url = URL(string: "https://www.example.com/")! receivedData = Data() let task = session.dataTask(with: url) task.resume() } // delegate methods func urlSession(_ session: URLSession, dataTask: URLSessionDataTask, didReceive response: URLResponse, completionHandler: @escaping (URLSession.ResponseDisposition) -> Void) { guard let response = response as? HTTPURLResponse, (200...299).contains(response.statusCode), let mimeType = response.mimeType, mimeType == "text/html" else { completionHandler(.cancel) return } completionHandler(.allow) } func urlSession(_ session: URLSession, dataTask: URLSessionDataTask, didReceive data: Data) { self.receivedData?.append(data) } func urlSession(_ session: URLSession, task: URLSessionTask, didCompleteWithError error: Error?) { DispatchQueue.main.async { self.loadButton.isEnabled = true if let error = error { handleClientError(error) } else if let receivedData = self.receivedData, let string = String(data: receivedData, encoding: .utf8) { self.webView.loadHTMLString(string, baseURL: task.currentRequest?.url) } } } The various delegate protocols offer methods beyond those shown in the above code, for handling authentication challenges, following redirects, and other special cases. Using a URL Session, in the URLSession discussion, describes the various callbacks that may occur during a transfer. ## See Also ### Essentials - [Analyzing HTTP traffic with Instruments](foundation/analyzing-http-traffic-with-instruments.md) - [URLSession](foundation/urlsession.md) - [URLSessionTask](foundation/urlsessiontask.md)