Final yr, iOS 15 got here with a really helpful function often called Live Text. You will have heard of the time period OCR (quick for Optical Character Recognition), which is the method for changing a picture of textual content right into a machine-readable textual content format. That is what Reside Textual content is about.
Reside Textual content is built-into the digital camera app and Pictures app. Should you haven’t tried out this function, merely open the Digital camera app. While you level the system’s digital camera at a picture of textual content, you will see that a Reside Textual content button on the lower-right nook. By tapping the button, iOS routinely captures the textual content for you. You’ll be able to then copy and paste it into different functions (e.g. Notes).
This can be a very highly effective and handy options for many customers. As a developer, wouldn’t or not it’s nice in case you can incorporate this Reside Textual content function in your individual app? In iOS 16, Apple launched the Reside Textual content API for builders to energy their apps with Reside Textual content. On this tutorial, let’s see methods to use the Reside Textual content API with SwiftUI.
Allow Reside Textual content Utilizing DataScannerViewController
Within the WWDC session about Capturing Machine-readable Codes and Text with VisionKit, Apple’s engineer confirmed the next diagram:
Textual content recognization will not be a brand new function on iOS 16. On older model of iOS, you should utilize APIs from the AVFoundation and Imaginative and prescient framework to detect and acknowledge textual content. Nevertheless, the implementation is kind of difficult, particularly for individuals who are new to iOS improvement.
Within the subsequent launch of iOS, all of the above is simplified to a brand new class known as DataScannerViewController in VisionKit. By utilizing this view controller, your app can routinely show a digital camera UI with the Reside Textual content functionality.
To make use of the category, you first import the VisionKit framework after which examine if the system helps the info scanner function:
|
DataScannerViewController.isSupported |
The Reside Textual content API solely helps gadgets launched in 2018 or newer with Neural engine. On prime of that, you additionally have to examine the supply to see if the consumer approves the usage of information scanner:
|
DataScannerViewController.isAvailable |
As soon as each checks come again with constructive outcomes, you might be prepared to begin the scanning. Right here is the pattern code to launch a digital camera with Reside Textual content:
|
let dataScanner = DataScannerViewController( recognizedDataTypes: [.text()], qualityLevel: .balanced, isHighlightingEnabled: true )
current(dataScanner, animated: true) strive? dataScanner.startScanning()
|
All you want is create an occasion of DataScannerViewController and specify the acknowledged information sorts. For textual content recognition, you move .textual content() as the info sort. As soon as the occasion is prepared, you’ll be able to current it and begin the scanning course of by calling the startScanning() methodology.
Working with DataScannerViewController in SwiftUI
The DataScannerViewController class now solely helps UIKit. For SwiftUI, it wants a bit of labor. We now have to undertake the UIViewControllerRepresentable protocol to make use of the category in SwiftUI initiatives. On this case, I create a brand new struct named DataScanner like this:
|
struct DataScanner: UIViewControllerRepresentable . . .
|
This struct accepts two binding variables: one for triggering the info scanning and the opposite is a binding for storing the scanned textual content.
|
@Binding var startScanning: Bool @Binding var scanText: String |
To efficiently undertake the UIViewControllerRepresentable protocol, we have to implement the next strategies:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 |
func makeUIViewController(context: Context) –> DataScannerViewController let controller = DataScannerViewController( recognizedDataTypes: [.text()], qualityLevel: .balanced, isHighlightingEnabled: true )
return controller
func updateUIViewController(_ uiViewController: DataScannerViewController, context: Context)
if startScanning strive? uiViewController.startScanning() else uiViewController.stopScanning()
|
Within the makeUIViewController methodology, we return an occasion of DataScannerViewController. For updateUIViewController, we begin (or cease) the scanning relying on the worth of startScanning.
To seize the scanned textual content, we now have to undertake the next methodology of DataScannerViewControllerDelegate:
|
func dataScanner(_ dataScanner: DataScannerViewController, didTapOn merchandise: RecognizedItem) . . .
|
The tactic known as when the consumer faucets the detected textual content, so we are going to implement it like this:
|
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 |
class Coordinator: NSObject, DataScannerViewControllerDelegate var mum or dad: DataScanner
init(_ mum or dad: DataScanner) self.mum or dad = mum or dad
func dataScanner(_ dataScanner: DataScannerViewController, didTapOn merchandise: RecognizedItem) swap merchandise case .textual content(let textual content): mum or dad.scanText = textual content.transcript default: break
func makeCoordinator() –> Coordinator Coordinator(self)
|
We examine the acknowledged merchandise and retailer the scanned textual content if any textual content is acknowledged. Lastly, insert this line of code within the makeUIViewController methodology to configure the delegate:
|
controller.delegate = context.coordinator |
This controller is now prepared to be used in SwiftUI views.
Capturing Textual content Utilizing DataScanner
For example, we are going to construct a textual content scanner app with a quite simple consumer interface. When the app is launched, it routinely shows the digital camera view for reside textual content. When textual content is detected, customers can faucet the textual content to seize it. The scanned textual content is displayed within the decrease a part of the display screen.
Assuming you’ve created a normal SwiftUI challenge, open ContentView.swift and the VisionKit framework.
Subsequent, declare a few state variables to manage the operation of the info scanner and the scanned textual content.
|
@State personal var startScanning = false @State personal var scanText = “” |
For the physique half, let’s replace the code like this:
|
VStack(spacing: 0) DataScanner(startScanning: $startScanning, scanText: $scanText) .body(top: 400)
Textual content(scanText) .body(minWidth: 0, maxWidth: .infinity, maxHeight: .infinity) .background(in: Rectangle()) .backgroundStyle(Shade(uiColor: .systemGray6))
.activity if DataScannerViewController.isSupported && DataScannerViewController.isAvailable startScanning.toggle()
|
We begin the info scanner when the app launches. However earlier than that, we name DataScannerViewController.isSupported and DataScannerViewController.isAvailable to make sure Reside Textual content is supported on the system.
The demo app is nearly prepared to check. Since Reside Textual content requires Digital camera entry, please bear in mind to go to the challenge configuration. Add the important thing Privateness – Digital camera Utilization Description within the Data.plist file and specify the rationale why your app must entry the system’s digital camera.
After the modifications, you’ll be able to deploy the app to an actual iOS system and check the Reside Textual content operate.
Aside from English, Reside Textual content additionally helps French, Italian, German, Spanish, Chinese language, Portuguese, Japanese, and Korean.
For the supply code of this demo challenge, you’ll be able to check it out on GitHub.
Observe: We’re updating our Mastering SwiftUI book for iOS 16. If you wish to begin studying SwiftUI, try the e-book here. You’ll obtain a free replace later this yr.
