react-native-rectangle-doc-scanner 1.13.0 → 1.14.0

package/docs/native-module-architecture.md

@@ -1,178 +0,0 @@
-# Native Module Architecture Plan
-
-This document lays out the roadmap for migrating the library from its current
-VisionCamera/OpenCV implementation to a fully native document scanner module
-that mirrors the behaviour of `react-native-document-scanner-plugin` while
-keeping support for custom React components and overlays.
-
-## Goals
-
-- Expose a `<DocScanner />` React view component that renders a native camera
-  preview.
-- Maintain the ability to layer custom React children (buttons, hints, headers)
-  on top of the preview.
-- Provide the detected polygon and stability information every frame, so the
-  existing Skia overlay and auto-capture logic continue working.
-- Support both automatic and manual capture flows; captures should resolve with
-  both cropped and original image metadata.
-- Keep the public TypeScript API as close as possible to the current version to
-  minimise breaking changes for consuming apps.
-
-## High-Level Architecture
-
-```
-┌──────────────────────────────────────────────────────────────────────┐
-│ React Native (TypeScript)                                            │
-│ ┌────────────────────────────┐   events / commands   ┌─────────────┐ │
-│ │ DocScanner.tsx (wrapper)  ├───────────────────────▶│ Native View │ │
-│ │ - Manages refs            │◀───────────────────────┤  Module     │ │
-│ │ - Handles props           │   layout callbacks     └─────────────┘ │
-│ │ - Renders Overlay + UI    │                                        │
-│ └────────────────────────────┘                                        │
-│                ▲                             ▲                       │
-│                │onRectangleDetect            │onPictureTaken         │
-└────────────────┴─────────────────────────────┴───────────────────────┘
-                 │                             │
-        ┌────────┴────────┐             ┌──────┴──────┐
-        │ iOS (Swift)      │             │ Android     │
-        │ DocScannerView   │             │ DocScannerView │
-        │ - AVCapture      │             │ - CameraX       │
-        │ - Vision / VNDetectRectangles │ │ - ML Kit / OpenCV│
-        │ - VNDocumentCamera for capture ││ - On-device cropping │
-        └──────────────────┘             └───────────────────────┘
-```
-
-### React Native Layer
-
-- `DocScanner.tsx`
-  - Forwards props to the native view via `requireNativeComponent`.
-  - Holds a ref to call `capture()`, `setTorchEnabled()`, etc.
-  - Converts native events into the existing `Point[]` / stability format.
-  - Continues rendering the Skia-based `Overlay` and any child components.
-- `index.ts`
-  - Exports the wrapper and new types.
-- TypeScript definitions for native events, capture result, and imperative
-  methods (`DocScannerHandle`).
-
-### iOS Implementation
-
-- Create `ios/DocScannerView.swift`
-  - Subclass of `UIView` hosting an `AVCaptureVideoPreviewLayer`.
-  - Uses `AVCaptureSession` with `AVCaptureVideoDataOutput` to stream frames.
-  - Runs `VNDetectRectanglesRequest` on a background queue for polygon
-    detection.
-  - Emits `onRectangleDetect` events (containing points, stability counter, and
-    frame size) via `RCTDirectEventBlock`.
-  - Implements `capture()` by calling `AVCapturePhotoOutput` and optionally
-    running `CIFilter` warps for perspective correction.
-  - Saves both original and cropped images, returning their URIs plus width /
-    height.
-
-- Create `ios/DocScannerViewManager.m`
-  - Registers the component as `RNRDocScannerView` (name to be decided).
-  - Exposes props: `overlayColor`, `detectionCountBeforeCapture`, `torchEnabled`,
-    `useBase64`, `quality`, etc.
-  - Exposes commands / methods: `capture`, `setTorchEnabled`, `reset`.
-
-- Create backing Swift files:
-  - `DocScannerFrameProcessor.swift` – handles rectangle detection and stability
-    scoring.
-  - `DocScannerCaptureCoordinator.swift` – manages photo output, perspective
-    correction, optional VisionKit integration.
-  - `DocScannerEventPayload.swift` – strongly typed event payloads.
-
-### Android Implementation
-
-- Create `android/src/main/java/.../DocScannerView.kt`
-  - Extends `FrameLayout`, hosts a `PreviewView` from CameraX.
-  - Sets up `ProcessCameraProvider` with `ImageAnalysis` for frame processing.
-  - Runs ML Kit Document Scanner API (or OpenCV fallback) to detect quads.
-  - Emits events through `@ReactProp` + `UIManagerModule` event dispatch.
-  - Maintains stability counter similar to current JS implementation.
-
-- Create `DocScannerViewManager.kt`
-  - Registers the view name (e.g. `RNRDocScannerView`).
-  - Maps props (`overlayColor`, `detectionCountBeforeCapture`, etc.).
-  - Implements `receiveCommand` to handle `capture`, `torch`, `reset`.
-
-- Create `DocScannerModule.kt`
-  - Exposes imperative methods if needed for non-view functionality (e.g. file
-    cleanup).
-
-- Provide utility classes:
-  - `ImageCropper.kt` – applies perspective transformations using OpenCV or
-    RenderScript.
-  - `PathUtils.kt` – file management for temporary images.
-  - `RectangleEvent.kt` – serialisable event payload.
-
-### Event Contract
-
-`onRectangleDetect` event shape (identical for iOS & Android):
-
-```ts
-type RectangleEventPayload = {
-  rectangleCoordinates: {
-    topLeft: Point;
-    topRight: Point;
-    bottomRight: Point;
-    bottomLeft: Point;
-  } | null;
-  stableCounter: number;
-  frameWidth: number;
-  frameHeight: number;
-};
-```
-
-`onPictureTaken` event shape:
-
-```ts
-type PictureEvent = {
-  croppedImage: string | null; // file:// or base64 depending on props
-  initialImage: string;
-  width: number;
-  height: number;
-};
-```
-
-### Props Mapping
-
-| Prop                          | Description                                         | Native handling                       |
-|-------------------------------|-----------------------------------------------------|---------------------------------------|
-| `overlayColor`               | Stroke colour for overlay (forwarded to events)     | Only metadata – actual overlay is JS  |
-| `detectionCountBeforeCapture`| Minimum stable frames before auto capture           | Native stability counter              |
-| `autoCapture`                | If `true`, native triggers capture automatically    | Native triggers `capture()` internally|
-| `enableTorch`                | Controls device torch                               | Maps to `AVCaptureDevice` / CameraX   |
-| `quality`                    | JPEG quality (0–100)                                | Controls compression in native capture|
-| `useBase64`                  | If `true`, return base64 strings instead of files   | Native encode                          |
-| `showGrid`, `gridColor`...   | Handled in JS overlay; no native changes required   | N/A                                   |
-
-### Migration Steps
-
-1. **Scaffold native module directories**
-   - Add `ios/` and `android/` folders with minimal React Native module wiring.
-
-2. **Implement iOS detection & capture**
-   - Set up camera session, rectangle detection, event dispatch.
-   - Support auto capture and manual capture commands.
-
-3. **Implement Android detection & capture**
-   - Mirror iOS logic with CameraX + ML Kit/OpenCV.
-
-4. **Update TypeScript wrapper**
-   - Replace VisionCamera/OpenCV logic with `requireNativeComponent`.
-   - Keep existing overlay and UI contract intact.
-
-5. **Testing & validation**
-   - Write example app update (outside this package) to verify events, overlay,
-    auto-capture.
-   - Ensure TypeScript typings align with native behaviour.
-
-6. **Documentation**
-   - Update README with new installation instructions (pods/gradle changes,
-    permissions, peer deps).
-   - Document new props & events, auto/manual capture behaviour.
-
-This plan focuses on creating a native module while preserving the custom
-overlay UX. Each platform will require significant native code, but the above
-structure provides the blueprint for implementation.
-

package/ios/RNRDocScannerModule.swift

@@ -1,49 +0,0 @@
-import Foundation
-import React
-
-@objc(RNRDocScannerModule)
-class RNRDocScannerModule: NSObject, RCTBridgeModule {
-  static func moduleName() -> String! {
-    "RNRDocScannerModule"
-  }
-
-  static func requiresMainQueueSetup() -> Bool {
-    true
-  }
-
-  @objc var bridge: RCTBridge?
-
-  @objc func capture(_ reactTag: NSNumber, resolver resolve: @escaping RCTPromiseResolveBlock, rejecter reject: @escaping RCTPromiseRejectBlock) {
-    bridge?.uiManager?.addUIBlock { _, viewRegistry in
-      guard let view = viewRegistry?[reactTag] as? RNRDocScannerView else {
-        reject(RNRDocScannerError.viewNotFound.code, RNRDocScannerError.viewNotFound.message, nil)
-        return
-      }
-
-      view.capture { result in
-        switch result {
-        case let .success(payload):
-          resolve([
-            "croppedImage": payload.croppedImage as Any,
-            "initialImage": payload.originalImage,
-            "width": payload.width,
-            "height": payload.height,
-          ])
-        case let .failure(error as RNRDocScannerError):
-          reject(error.code, error.message, error)
-        case let .failure(error):
-          reject("capture_failed", error.localizedDescription, error)
-        }
-      }
-    }
-  }
-
-  @objc func reset(_ reactTag: NSNumber) {
-    bridge?.uiManager?.addUIBlock { _, viewRegistry in
-      guard let view = viewRegistry?[reactTag] as? RNRDocScannerView else {
-        return
-      }
-      view.resetStability()
-    }
-  }
-}