From efa152cc9811ac275beefc3dc3f3a4a48f2c95b1 Mon Sep 17 00:00:00 2001
From: Jonathan Palma <tanpalma04@gmail.com>
Date: Thu, 2 Jul 2020 00:58:33 -0600
Subject: [PATCH] feat(android): add tesseract support [UNSAFE]

---
 android/build.gradle                          |   1 +
 .../com/reactlibrary/TesseractOcrModule.java  | 119 +++++++++++++++++-
 index.js                                      |   5 +-
 3 files changed, 121 insertions(+), 4 deletions(-)

diff --git a/android/build.gradle b/android/build.gradle
index 5dd382b..885c6e5 100644
--- a/android/build.gradle
+++ b/android/build.gradle
@@ -73,6 +73,7 @@ repositories {
 dependencies {
     //noinspection GradleDynamicVersion
     implementation 'com.facebook.react:react-native:+'  // From node_modules
+    implementation 'com.rmtheis:tess-two:9.1.0'
 }
 
 def configureReactNativePom(def pom) {
diff --git a/android/src/main/java/com/reactlibrary/TesseractOcrModule.java b/android/src/main/java/com/reactlibrary/TesseractOcrModule.java
index 3e4274a..0923b34 100644
--- a/android/src/main/java/com/reactlibrary/TesseractOcrModule.java
+++ b/android/src/main/java/com/reactlibrary/TesseractOcrModule.java
@@ -1,5 +1,9 @@
 package com.reactlibrary;
 
+import android.os.Handler;
+import android.support.annotation.Nullable;
+import android.util.Log;
+
 import com.facebook.react.bridge.ReactApplicationContext;
 import com.facebook.react.bridge.ReactContextBaseJavaModule;
 import com.facebook.react.bridge.ReactMethod;
@@ -8,10 +12,20 @@
 public class TesseractOcrModule extends ReactContextBaseJavaModule {
 
     private final ReactApplicationContext reactContext;
+    private Handler handler = new Handler();
+    private TessBaseAPI tesseract;
+
+    private static final String KEY_ALLOW_LIST = "allowlist";
+    private static final String KEY_DENY_LIST = "denylist";
+    private static final String DATA_PATH = Environment.getExternalStorageDirectory().toString() + File.separator;
+	  private static final String DATA_FILES_DIRECTORY = "tessdata";
 
     public TesseractOcrModule(ReactApplicationContext reactContext) {
         super(reactContext);
         this.reactContext = reactContext;
+        if (!this.DATA_PATH.contains(reactContext.getPackageName())) {
+            this.DATA_PATH += reactContext.getPackageName() + File.separator;
+        }
     }
 
     @Override
@@ -20,8 +34,107 @@ public String getName() {
     }
 
     @ReactMethod
-    public void sampleMethod(String stringArgument, int numberArgument, Callback callback) {
-        // TODO: Implement some actually useful functionality
-        callback.invoke("Received numberArgument: " + numberArgument + " stringArgument: " + stringArgument);
+    public void stop(Promise promise) {
+		    Log.d(getName(), "stop");
+        try {
+            tesseract.stop();
+            tesseract.end();
+            promise.resolve("Recognition was canceled");
+        } catch (Exception e) {
+            Log.e(getName(), e.toString());
+            promise.reject("An error occurred when trying to stop recognition", e.toString());
+        }
+    }
+
+    @SuppressLint("StaticFieldLeak")
+    @ReactMethod
+    public void recognize(String path, String lang, @Nullable ReadableMap options, final Callback callback, final Promise promise) {
+        Log.d(getName(), "recognize");
+
+        try {
+            prepareTesseract();
+            BitmapFactory.Options options = new BitmapFactory.Options();
+            Bitmap bitmap = BitmapFactory.decodeFile(path, options);
+
+            new Thread() {
+                @Override
+                public void run() {
+                    tesseract = new TessBaseAPI(new TessBaseAPI.ProgressNotifier() {
+                        @Override
+                        public void onProgressValues(TessBaseAPI.ProgressValues progressValues) {
+                            Log.d("progress " + progressValues.getPercent().toString());
+                            callback(progressValues.getPercent());
+                        }
+                    });
+                    tesseract.init(DATA_PATH, lang);
+                    tesseract.setImage(bitmap);
+                    tesseract.getHOCRText(0);
+
+                    String recognizedText = tesseract.getUTF8Text();
+
+                    tesseract.end();
+                    promise.resolve(recognizedText);
+                }
+
+            }.start();
+
+        } catch (Exception e) {
+            Log.e(getName(), e.getMessage());
+            promise.reject("Could not recognize text", e.toString());
+        }
+    }
+
+    private void prepareDirectory(String path) {
+        Log.d(getName(), "prepare directory");
+
+        File dir = new File(path);
+        if (!dir.exists()) {
+            if (!dir.mkdirs()) {
+                Log.e(getName(), "Could not create directory (" + path + "), please make sure the app is allowed to write to external storage.");
+            }
+        } else {
+            Log.d(getName(), "Created directory " + path);
+        }
+    }
+
+    private void prepareTesseract() {
+        Log.d(getName(), "prepare tesseract environment");
+
+        try {
+            prepareDirectory(DATA_PATH + DATA_FILES_DIRECTORY);
+        } catch (Exception e) {
+            e.printStackTrace();
+        }
+
+        copyTessDataFiles(DATA_FILES_DIRECTORY);
+    }
+
+    private void copyTessDataFiles(String path) {
+        Log.d(getName(), "copy tesseract data files");
+        try {
+            String fileList[] = reactContext.getAssets().list(path);
+
+            for (String fileName : fileList) {
+                String oldPath = path + File.separator + fileName;
+                String newPath = DATA_PATH + oldPath;
+
+                if (!(new File(newPath)).exists()) {
+                    InputStream in = reactContext.getAssets().open(oldPath);
+                    OutputStream out = new FileOutputStream(newPath);
+                    byte[] buf = new byte[1024];
+                    int len;
+
+                    while ((len = in.read(buf)) > 0) {
+                        out.write(buf, 0, len);
+                    }
+                    in.close();
+                    out.close();
+
+                    Log.d(getName(), "Copied " + oldPath + " to " + newPath);
+                }
+            }
+        } catch (IOException e) {
+            Log.e(getName(), "Unable to copy files to tessdata", e.toString());
+        }
     }
 }
diff --git a/index.js b/index.js
index 7d46aa0..49ec8eb 100644
--- a/index.js
+++ b/index.js
@@ -1,5 +1,8 @@
-import { NativeModules } from 'react-native';
+import { NativeModules } from "react-native";
 
 const { TesseractOcr } = NativeModules;
 
+TesseractOcr.recognize = (path, lang, { onProgress, ...rest }) =>
+  TesseractOcr.recognize(path, lang, rest, onProgress);
+
 export default TesseractOcr;