How to bridge platform-specific types in Swift & SwiftUI

Apr 25, 2022 · Follow on Twitter and Mastodon swiftui swift multi-platform

SwiftUI’s multi-platform support makes it easy to develop apps for iOS, macOS, tvOS and watchOS. But how do you handle types that differ between platforms? Let’s take a look.

Problem description

Lets say that we fetch the following list item over a REST-based API, with the intention to list it in a multi-platform SwiftUI app:

struct ListItem: Codable {

    let title: String
    let created: Date
    let imageData: Data
}

The type itself is platform-agnostic, but to do anything with the image, we need to convert it to UIImage for UIKit, NSImage for AppKit and Image for SwiftUI.

Let’s take a look at how we can provide a displayable image in a platform-agnostic way.

UIKit & AppKit

To support UIKit and AppKit, we can add an image extension to ListItem that maps the imageData value to either a UIImage or an NSImage:

import SwiftUI

#if canImport(UIKit)
extension ListItem {

    var image: UIImage? {
        .init(data: imageData)
    }
}
#elseif canImport(AppKit)
extension ListItem {

    var image: NSImage? {
        .init(data: imageData)
    }
}
#endif

To avoid having to use #if checks everywhere in the code, I actually first prefer to define a platform-agnostic image typealias like this:

import SwiftUI

#if canImport(UIKit)
public typealias ImageRepresentable = UIImage
#elseif canImport(AppKit)
public typealias ImageRepresentable = NSImage
#endif

Since both UIImage and NSImage have a data-based initializer, this ImageRepresentable typealias now lets us rewrite the image property like this:

extension ListItem {

    var nativeImage: ImageRepresentable {
        ImageRepresentable(data: imageData)
    }
}

We can then add any new capabilities that we need to UIImage and NSImage, to have a platform-agnostic image type that works in the same way across all platforms.

SwiftUI

To support SwiftUI, we can extend Image to make it easier to initialize it with this new type:

import SwiftUI

extension Image {
    
    init(_ image: ImageRepresentable) {
        #if canImport(UIKit)
        self.init(uiImage: image)
        #elseif canImport(Cocoa)
        self.init(nsImage: image)
        #endif
    }
}

We can now extend ListItem with a SwiftUI image without having to do any #if checks:

extension ListItem {

    var image: Image { 
        .init(nativeImage) 
    }
}

Extending the platform-agnostic image type

This was easy to achieve, since UIImage and NSImage both had a Data-based initializer, but how about when they don’t share the same APIs?

For instance, consider how UIImage has a jpegData(compressionQuality:) function that NSImage lacks. We can then fill in the gaps by implementing missing functionality.

We can implement jpegData for NSImage by first defining a cgImage property:

#if canImport(Cocoa)
extension NSImage {
    
    var cgImage: CGImage? {
        cgImage(forProposedRect: nil, context: nil, hints: nil)
    }
}
#endif

We can then use this function to define a jpegData function:

#if canImport(Cocoa)
extension NSImage {
 
    func jpegData(compressionQuality: CGFloat) -> Data? {
        guard let image = cgImage else { return nil }
        let bitmap = NSBitmapImageRep(cgImage: image)
        return bitmap.representation(using: .jpeg, properties: [.compressionFactor: compressionQuality])
    }
}
#endif

Since both UIImage and NSImage now have a jpegData function with the same signature, you can extend ImageRepresentable by building upon the shared functionality:

extension ImageRepresentable {

    func compressedForSharing() -> Self? {
        jpegData(compressionQuality: 0.7)
    }
}

Since both types define the same API, we don’t need to add any #if checks. This keeps the rest of our source code clean and less error-prone.

Discussions & More

Please share any ideas, feedback or comments you may have in the Disqus section below, or by replying to this tweet or this toot.

If you found this text interesting, make sure to follow me on Twitter and Mastodon for more content like this, and to be notified when new content is published.

If you like & want to support my work, please consider sponsoring me on GitHub Sponsors.