Swift Package Manage使用及自定义

参考：

• Swift Package Manager 使用
• iOS包依赖管理工具（五）：Swift Package Manager（SPM）自定义篇
• SwiftUI中文入门
• SwiftUI_learning01
• swiftui-state-property-binding-stateobject-observedobject-environmentobject-學習筆記

前言

最近在接入亚马逊人脸识别时，亚马逊提供的最新SDK是使用Swift Package Manage管理的，而项目里使用的其他三方SDK是使用cocoa pods进行管理，一开始以为两个不能兼容，但是经过实际测试，是可以正常一起使用

普通项目或Pod项目中引入Swift Package Manager管理的项目

以接入amplify-swift git地址：https://github.com/aws-amplify/amplify-swift

以接入amplify-ui-swift-liveness git地址：https://github.com/aws-amplify/amplify-ui-swift-liveness

---

UIKit里引入SwiftUI框架

由于引入的AmplifyUILiveness是使用的SwiftUI，我们的主题项目使用的是标准UIKit，所以需要再普通页面引入需要进行包装一成

参考：

• SwiftUI 与 UIKit 混合开发 使用UIHostingController进行包装

  import UIKit
  import SwiftUI
  class ViewController: UIViewController {
    override func viewDidLoad() {
        super.viewDidLoad()
        let btn = UIButton(frame: CGRect(x: 100, y: 100, width: 60, height: 40))
        btn.setTitle("showTest", for: .normal)
        btn.setTitleColor(.black, for: .normal)
        btn.addTarget(self, action: #selector(showMyTestSwiftUI), for: .touchUpInside)
        self.view.addSubview(btn)
    }
      
    @objc func showMyTestSwiftUI() {
        let vc = UIHostingController(rootView: Text("Text"))
        self.navigationController?.pushViewController(vc, animated: true)
    }
  }
  

  ---

  自定义Swift Package Manager管理的项目

  由于我们的主项目需要支持到iOS12，但SwiftUI最低需要iOS14,所以需要把AmplifyUILiveness下载好源码改成普通UIKit的框架

本地调试SPM项目： 参考：iOS包依赖管理工具（五）：Swift Package Manager（SPM）自定义篇

1、创建Package项目：Xcode -> File -> New -> Project 2、将自己创建的或网上下载的SPM的Package项目拖拽到测试项目，一定要将生成的包含Package.swift的那个文件夹拖拽到项目里 2.2 拖拽到项目里后显示的是这样 2.3 在Targets -> xxx -> Gegeral -> Frameworks, Libraries, and Embeded Content -> + 2.4 选中对应的库，然后选择Add到项目里 2.4、以接入的项目里为例，在项目中使用

import UIKit
import FaceLiveness
import SwiftUI
import AWSPluginsCore
import Amplify

class ViewController: UIViewController {

    override func viewDidLoad() {
        super.viewDidLoad()
        // Do any additional setup after loading the view.
        
        let btn = UIButton(type: .custom)
        btn.setTitle("test", for: .normal)
        btn.backgroundColor = .lightGray
        btn.setTitleColor(.systemBlue, for: .normal)
        btn.addTarget(self, action: #selector(testClick), for: .touchUpInside)
        btn.frame = .init(x: 100, y: 100, width: 100, height: 50)
        self.view.addSubview(btn)
    }

    @objc func testClick() {
        let vc = UIHostingController(rootView: MyFaceLiveCheckView())
        self.navigationController?.pushViewController(vc, animated: true)
    }
    
}



struct MyFaceLiveCheckView: View {
    @State private var isPresentingLiveness = true
        
      var body: some View {
        FaceLivenessDetectorView(
            sessionID: "xxx",
          credentialsProvider: MyCredentialsProvider(),
          region: "us-east-1",
          isPresented: $isPresentingLiveness,
          onCompletion: { result in
            switch result {
            case .success:
                // 检测结果
//                checkResult()
                break
            case .failure(let error):
                // 本地活体SDK报错
                print(error)
                break
            }
          }
        )
      }
    
}
struct MyCredentialsProvider: AWSCredentialsProvider {
    
    func fetchAWSCredentials() async throws -> AWSCredentials {
        return MyCredentials()
    }
    
    struct MyCredentials: AWSCredentials {
        var accessKeyId: String = "xxx"
        var secretAccessKey: String = "xxxx"
    }
}

---

核心操作：FaceLivenessDetectionViewModel 等待页面：GetReadyPageView -> CameraPreviewView -> previewCaptureSession?.startSession 开始 FaceLivenessDetectorView -> onBegin -> displayState = .displayingLiveness -> _FaceLivenessDetectionView 人脸扫描页面： _FaceLivenessDetectionView -> makeUIViewController -> _LivenessViewController -> viewDidAppear -> setupAVLayer -> viewModel.startCamera -> captureSession.startSession -> videoOutput.setSampleBufferDelegate -> 系统delegate回调 -> captureOutput -> FaceDetectorShortRange.detectFaces -> prediction(人脸数) -> detectionResultHandler?.process -> .singleFace -> 一个人脸且符合要求的人脸 -> 初始化配置 -> initializeLivenessStream -> livenessService?.initializeLivenessStream -> 链接aws服务器 -> websocket.open(url: xx) 有新视频流后： detectionResultHandler?.process -> .recording -> drawOval -> sendInitialFaceDetectedEvent

摄像头准备好后开始初始化Stream FaceLivenessDetectorView.init -> LivenessCaptureSession -> OutputSampleBufferCapturer -> FaceLivenessDetectionViewModel OutputSampleBufferCapturer -> faceDetector.detectFaces -> detectionResultHandler?.process FaceLivenessDetectionViewModel -> FaceDetectionResultHandler -> process -> initializeLivenessStream

视频流核心： CameraPreviewViewModel