hotfix: Extend Tesseract instead of Tesseract1

buildSrc/src/main/kotlin/com.iqser.red.service.java-conventions.gradle.kts

@@ -12,10 +12,6 @@ group = "com.knecon.fforesight.service"
 java.sourceCompatibility = JavaVersion.VERSION_17
 java.targetCompatibility = JavaVersion.VERSION_17
 
-pmd {
-    isConsoleOutput = true
-}
-
 tasks.pmdMain {
     pmd.ruleSetFiles = files("${rootDir}/config/pmd/pmd.xml")
 }

config/pmd/pmd.xml

@@ -9,11 +9,12 @@
     </description>
 
     <rule ref="category/java/errorprone.xml">
+        <exclude name="DataflowAnomalyAnalysis"/>
         <exclude name="MissingSerialVersionUID"/>
+        <exclude name="NullAssignment"/>
         <exclude name="AvoidLiteralsInIfCondition"/>
         <exclude name="AvoidDuplicateLiterals"/>
-        <exclude name="NullAssignment"/>
-        <exclude name="AssignmentInOperand"/>
+        <exclude name="AvoidFieldNameMatchingMethodName"/>
     </rule>
 
-</ruleset>
+</ruleset>

config/pmd/test_pmd.xml

@@ -10,13 +10,14 @@
 
 
     <rule ref="category/java/errorprone.xml">
+        <exclude name="DataflowAnomalyAnalysis"/>
         <exclude name="MissingSerialVersionUID"/>
+        <exclude name="NullAssignment"/>
         <exclude name="AvoidLiteralsInIfCondition"/>
         <exclude name="AvoidDuplicateLiterals"/>
-        <exclude name="NullAssignment"/>
-        <exclude name="AssignmentInOperand"/>
+        <exclude name="AvoidFieldNameMatchingMethodName"/>
+        <exclude name="AvoidFieldNameMatchingTypeName"/>
         <exclude name="TestClassWithoutTestCases"/>
-        <exclude name="BeanMembersShouldSerialize"/>
     </rule>
 
-</ruleset>
+</ruleset>

ocr-service-v1/ocr-service-processor/build.gradle.kts

@@ -24,6 +24,6 @@ dependencies {
     api("io.github.karols:hocr4j:0.2.0")
     api("com.amazonaws:aws-java-sdk-kms:1.12.440")
     api("com.google.guava:guava:31.1-jre")
-    api("com.iqser.red.commons:pdftron-logic-commons:2.23.0")
+    api("com.iqser.red.commons:pdftron-logic-commons:2.20.0")
     testImplementation("org.junit.jupiter:junit-jupiter:5.8.1")
 }

ocr-service-v1/ocr-service-processor/src/main/java/com/knecon/fforesight/service/ocr/processor/service/OCRService.java

@@ -28,8 +28,6 @@ import com.knecon.fforesight.service.ocr.processor.service.scriptdetection.FontS
 import com.knecon.fforesight.service.ocr.processor.service.threads.OCRThread;
 import com.knecon.fforesight.service.ocr.processor.settings.OcrServiceSettings;
 
-import io.micrometer.observation.ObservationRegistry;
-import io.micrometer.observation.annotation.Observed;
 import lombok.AccessLevel;
 import lombok.RequiredArgsConstructor;
 import lombok.SneakyThrows;
@@ -50,7 +48,6 @@ public class OCRService {
     OcrResultWriter ocrResultWriter;
     GhostScriptService ghostScriptService;
     FontStyleDetector boldDetector;
-    ObservationRegistry registry;
 
 
     /**
@@ -62,13 +59,12 @@ public class OCRService {
      * @param fileId    Id of file
      * @param out       OutputStream where to write to
      */
-    @Observed(name = "OCRService", contextualName = "run-ocr-on-document")
     @SneakyThrows
     public void runOcrOnDocument(String dossierId, String fileId, OutputStream out) {
 
         try (InputStream fileStream = removeWatermarkIfEnabled(dossierId, fileId); ByteArrayOutputStream transferOutputStream = new ByteArrayOutputStream()) {
 
-            invisibleElementRemovalService.removeInvisibleElements(fileStream, transferOutputStream, false, false);
+            invisibleElementRemovalService.removeInvisibleElements(fileStream, transferOutputStream, false);
 
             try (InputStream transferInputStream = new ByteArrayInputStream(transferOutputStream.toByteArray())) {
                 log.info("Starting OCR for file {}", fileId);

ocr-service-v1/ocr-service-processor/src/main/java/com/knecon/fforesight/service/ocr/processor/service/threads/GhostScriptOutputHandler.java

@@ -109,6 +109,7 @@ public class GhostScriptOutputHandler extends Thread {
         if (imageFile == null) {
             throw new IllegalArgumentException(String.format("Page number %d does not exist in this thread. It only has pagenumbers %s", pageNumber, pagesToProcess.keySet()));
         }
+        assert new File(imageFile.absoluteFilePath()).isFile();
         renderedPageImageFileOutput.add(imageFile);
     }
 

ocr-service-v1/ocr-service-processor/src/main/java/com/knecon/fforesight/service/ocr/processor/service/threads/ImageProcessingThread.java

@@ -198,10 +198,8 @@ public class ImageProcessingThread extends Thread {
             grayScale = pix;
         } else if (pix.d == 32) {
             grayScale = Leptonica1.pixConvertRGBToGrayFast(pix);
-            LeptUtils.disposePix(pix);
         } else if (pix.d == 1) {
             grayScale = Leptonica1.pixConvert1To8(null, pix, (byte) 0, (byte) 255);
-            LeptUtils.disposePix(pix);
         } else {
             throw new UnsupportedOperationException(String.format("Unknown pix format with bpp of %d", pix.d));
         }
@@ -210,27 +208,29 @@ public class ImageProcessingThread extends Thread {
         float targetFactor = targetDpi / imageDpi;
         if (targetFactor > 2.1) {
             scaledUp = Leptonica1.pixScaleGray4xLI(grayScale);
-            LeptUtils.disposePix(grayScale);
         } else if (targetFactor > 1.1) {
             scaledUp = Leptonica1.pixScaleGray2xLI(grayScale);
-            LeptUtils.disposePix(grayScale);
         } else {
             scaledUp = grayScale;
         }
 
         // remove noise and prep for Otsu
         gaussian = Leptonica1.pixConvolve(scaledUp, gaussianKernel, 8, 1);
-        LeptUtils.disposePix(scaledUp);
 
         // Threshold to binary
         if (pix.w < 100 || pix.h < 100) {
             binarized = Leptonica1.pixThresholdToBinary(gaussian, 170);
         } else {
             binarized = Leptonica1.pixOtsuThreshOnBackgroundNorm(gaussian, null, 50, 50, 165, 10, 100, 5, 5, 0.2f, null);
+
             if (binarized == null) { // Sometimes Otsu just fails, then we binarize directly
                 binarized = Leptonica1.pixThresholdToBinary(gaussian, 170);
             }
         }
+
+        LeptUtils.disposePix(pix);
+        LeptUtils.disposePix(grayScale);
+        LeptUtils.disposePix(scaledUp);
         LeptUtils.disposePix(gaussian);
 
         return binarized;

ocr-service-v1/ocr-service-processor/src/main/java/com/knecon/fforesight/service/ocr/processor/utils/Tesseract2.java

@@ -1,5 +1,25 @@
 package com.knecon.fforesight.service.ocr.processor.utils;
 
+import static net.sourceforge.tess4j.ITessAPI.TRUE;
+import static net.sourceforge.tess4j.TessAPI1.TessBaseAPIDelete;
+import static net.sourceforge.tess4j.TessAPI1.TessBaseAPIEnd;
+import static net.sourceforge.tess4j.TessAPI1.TessBaseAPIGetIterator;
+import static net.sourceforge.tess4j.TessAPI1.TessBaseAPIGetStringVariable;
+import static net.sourceforge.tess4j.TessAPI1.TessBaseAPIMeanTextConf;
+import static net.sourceforge.tess4j.TessAPI1.TessBaseAPIProcessPage;
+import static net.sourceforge.tess4j.TessAPI1.TessDeleteResultRenderer;
+import static net.sourceforge.tess4j.TessAPI1.TessHOcrRendererCreate;
+import static net.sourceforge.tess4j.TessAPI1.TessPageIteratorBegin;
+import static net.sourceforge.tess4j.TessAPI1.TessPageIteratorBoundingBox;
+import static net.sourceforge.tess4j.TessAPI1.TessPageIteratorNext;
+import static net.sourceforge.tess4j.TessAPI1.TessResultIteratorConfidence;
+import static net.sourceforge.tess4j.TessAPI1.TessResultIteratorDelete;
+import static net.sourceforge.tess4j.TessAPI1.TessResultIteratorGetPageIterator;
+import static net.sourceforge.tess4j.TessAPI1.TessResultIteratorGetUTF8Text;
+import static net.sourceforge.tess4j.TessAPI1.TessResultRendererBeginDocument;
+import static net.sourceforge.tess4j.TessAPI1.TessResultRendererEndDocument;
+import static net.sourceforge.tess4j.TessAPI1.TessResultRendererInsert;
+
 import java.awt.Rectangle;
 import java.nio.IntBuffer;
 import java.util.ArrayList;
@@ -9,20 +29,19 @@ import com.sun.jna.Pointer;
 
 import lombok.extern.slf4j.Slf4j;
 import net.sourceforge.lept4j.Pix;
+import net.sourceforge.tess4j.ITessAPI;
 import net.sourceforge.tess4j.OCRResult;
 import net.sourceforge.tess4j.TessAPI1;
-import net.sourceforge.tess4j.Tesseract1;
+import net.sourceforge.tess4j.Tesseract;
 import net.sourceforge.tess4j.TesseractException;
 import net.sourceforge.tess4j.Word;
 
 @Slf4j
 /**
  * Overriden version only so I can use Tesseract1 with Pixs instead of BufferedImages. All Functions are copied and then the BufferedImage -> Pix conversion deleted.
- */
-public class Tesseract2 extends Tesseract1 {
+ */ public class Tesseract2 extends Tesseract {
 
-
-    private int createDocuments(Pix pix, String filename, TessResultRenderer renderer) {
+    private int createDocuments(Pix pix, String filename, ITessAPI.TessResultRenderer renderer) {
 
         String title = TessBaseAPIGetStringVariable(getHandle(), DOCUMENT_TITLE);
         TessResultRendererBeginDocument(renderer, title);
@@ -62,7 +81,7 @@ public class Tesseract2 extends Tesseract1 {
         try {
             for (int i = 0; i < pixs.length; i++) {
                 try {
-                    TessResultRenderer renderer = createRenderers(outputbases[i], formats);
+                    ITessAPI.TessResultRenderer renderer = createRenderers(outputbases[i], formats);
                     int meanTextConfidence = createDocuments(pixs[i], filenames[i], renderer);
                     TessDeleteResultRenderer(renderer);
                     List<Word> words = meanTextConfidence > 0 ? getRecognizedWords(pageIteratorLevel) : new ArrayList<Word>();
@@ -85,8 +104,8 @@ public class Tesseract2 extends Tesseract1 {
         List<Word> words = new ArrayList<>();
 
         try {
-            TessResultIterator ri = TessBaseAPIGetIterator(getHandle());
-            TessPageIterator pi = TessResultIteratorGetPageIterator(ri);
+            ITessAPI.TessResultIterator ri = TessBaseAPIGetIterator(getHandle());
+            ITessAPI.TessPageIterator pi = TessResultIteratorGetPageIterator(ri);
             TessPageIteratorBegin(pi);
 
             do {
@@ -119,9 +138,9 @@ public class Tesseract2 extends Tesseract1 {
     }
 
 
-    private TessResultRenderer createRenderers(String outputbase, List<RenderedFormat> formats) {
+    private ITessAPI.TessResultRenderer createRenderers(String outputbase, List<RenderedFormat> formats) {
 
-        TessResultRenderer renderer = null;
+        ITessAPI.TessResultRenderer renderer = null;
 
         for (RenderedFormat format : formats) {
             switch (format) {
@@ -138,4 +157,12 @@ public class Tesseract2 extends Tesseract1 {
         return renderer;
     }
 
+
+    @Override
+    protected void dispose() {
+
+        TessBaseAPIEnd(getHandle());
+        TessBaseAPIDelete(getHandle());
+    }
+
 }

ocr-service-v1/ocr-service-server/src/main/java/com/knecon/fforesight/service/ocr/v1/server/queue/OcrMessageReceiver.java

@@ -21,7 +21,6 @@ import com.knecon.fforesight.service.ocr.v1.api.model.DocumentRequest;
 import com.iqser.red.service.persistence.service.v1.api.shared.model.dossiertemplate.dossier.file.FileErrorInfo;
 
 import feign.FeignException;
-import io.micrometer.observation.annotation.Observed;
 import lombok.AccessLevel;
 import lombok.RequiredArgsConstructor;
 import lombok.experimental.FieldDefaults;

ocr-service-v1/ocr-service-server/src/main/resources/vcpkg.json

@@ -6,7 +6,7 @@
   "overrides": [
     {
       "name": "tesseract",
-      "version": "5.3.3"
+      "version": "5.3.2"
     },
     {
       "name": "leptonica",

ocr-service-v1/ocr-service-server/src/test/java/com/knecon/fforesight/service/ocr/v1/server/OcrServiceIntegrationTest.java

@@ -64,7 +64,7 @@ public class OcrServiceIntegrationTest extends AbstractTest {
     @SneakyThrows
     public void testOcr() {
 
-        String text = testOCR("files/UNAPPROVED_VV-331155 (1).pdf");
+        String text = testOCR("files/402Study.pdf");
     }