clari-added-section-chunk: change structure object from paragraph/headline to sections

* add sections without tables

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/LayoutParsingPipeline.java

@@ -163,7 +163,7 @@ public class LayoutParsingPipeline {
         layoutParsingStorageService.storeSimplifiedText(layoutParsingRequest, simplifiedSectionTextService.toSimplifiedText(documentGraph));
         layoutParsingStorageService.storeViewerDocument(layoutParsingRequest, viewerDocumentFile);
 
-        if (layoutParsingRequest.layoutParsingType().equals(LayoutParsingType.CLARIFYND)) {
+        if (layoutParsingRequest.researchDocumentStorageId() != null) {
             log.info("Building research document data for {}", layoutParsingRequest.identifier());
             var researchDocumentData = TaasDocumentDataMapper.fromDocument(documentGraph);
             layoutParsingStorageService.storeResearchDocumentData(layoutParsingRequest, researchDocumentData);

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/text/TextPageBlock.java

@@ -1,7 +1,5 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.text;
 
-import static com.knecon.fforesight.service.layoutparser.processor.utils.ParsingConstants.NEW_LINE_TEXT_HEIGHT_PERCENTAGE;
-
 import java.awt.geom.Rectangle2D;
 import java.util.ArrayList;
 import java.util.List;
@@ -210,7 +208,7 @@ public class TextPageBlock extends AbstractPageBlock {
         TextPositionSequence previous = null;
         for (TextPositionSequence word : sequences) {
             if (previous != null) {
-                if (Math.abs(previous.getMaxYDirAdj() - word.getMaxYDirAdj()) > word.getTextHeight() * NEW_LINE_TEXT_HEIGHT_PERCENTAGE) {
+                if (Math.abs(previous.getMaxYDirAdj() - word.getMaxYDirAdj()) > word.getTextHeight()) {
                     sb.append('\n');
                 } else {
                     sb.append(' ');
@@ -230,7 +228,7 @@ public class TextPageBlock extends AbstractPageBlock {
         TextPositionSequence previous = null;
         for (TextPositionSequence word : sequences) {
             if (previous != null) {
-                if (word.getMaxYDirAdj() - previous.getMaxYDirAdj() > word.getTextHeight() * NEW_LINE_TEXT_HEIGHT_PERCENTAGE) {
+                if (word.getMaxYDirAdj() - previous.getMaxYDirAdj() > word.getTextHeight()) {
                     numberOfLines++;
                 }
             }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/factory/SearchTextWithTextPositionFactory.java

@@ -1,7 +1,5 @@
 package com.knecon.fforesight.service.layoutparser.processor.services.factory;
 
-import static com.knecon.fforesight.service.layoutparser.processor.utils.ParsingConstants.NEW_LINE_TEXT_HEIGHT_PERCENTAGE;
-
 import java.awt.geom.AffineTransform;
 import java.awt.geom.Rectangle2D;
 import java.util.Collection;
@@ -69,6 +67,7 @@ public class SearchTextWithTextPositionFactory {
             ++context.stringIdx;
         }
 
+
         List<Rectangle2D> positions = sequences.stream()
                 .map(TextPositionSequence::getTextPositions)
                 .flatMap(Collection::stream)
@@ -162,7 +161,7 @@ public class SearchTextWithTextPositionFactory {
         }
 
         double deltaY = Math.abs(currentPosition.getYDirAdj() - previousPosition.getYDirAdj());
-        return deltaY >= currentPosition.getHeightDir() * NEW_LINE_TEXT_HEIGHT_PERCENTAGE;
+        return deltaY >= currentPosition.getHeightDir();
     }
 
 
@@ -192,9 +191,9 @@ public class SearchTextWithTextPositionFactory {
 
         float textHeight = sequence.getTextHeight() + HEIGHT_PADDING;
         Rectangle2D rectangle2D = new Rectangle2D.Double(textPosition.getXDirAdj(),
-                                                         textPosition.getYDirAdj() - textHeight,
-                                                         textPosition.getWidthDirAdj(),
-                                                         textHeight + HEIGHT_PADDING);
+                textPosition.getYDirAdj() - textHeight,
+                textPosition.getWidthDirAdj(),
+                textHeight + HEIGHT_PADDING);
 
         AffineTransform transform = new AffineTransform();
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/mapper/TaasDocumentDataMapper.java

@@ -16,6 +16,7 @@ import com.knecon.fforesight.service.layoutparser.internal.api.data.taas.TableDa
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.Boundary;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Page;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Section;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SemanticNode;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Table;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.TableCell;
@@ -28,7 +29,8 @@ public class TaasDocumentDataMapper {
         AtomicInteger structureObjectNumber = new AtomicInteger();
         List<StructureObject> structureObjects = document.streamAllSubNodes()
                 .filter(node -> !node.getType().equals(NodeType.TABLE_CELL))
-                .filter(node -> !node.getType().equals(NodeType.SECTION))
+                .filter(node -> !node.getType().equals(NodeType.SUPER_SECTION))
+                .filter(node -> !(node.getType().equals(NodeType.SECTION) && ((Section) node).hasTables()))
                 .map(node -> {
                     if (node.getType().equals(NodeType.TABLE)) {
                         return TaasDocumentDataMapper.fromTableWithTableData((Table) node, structureObjectNumber.getAndIncrement());

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/parsing/LegacyPDFStreamEngine.java

@@ -82,7 +82,6 @@ public class LegacyPDFStreamEngine extends PDFStreamEngine {
 
     private int pageRotation;
     private PDRectangle pageSize;
-    private Matrix translateMatrix;
     private final GlyphList glyphList;
     private final Map<COSDictionary, Float> fontHeightMap = new WeakHashMap<COSDictionary, Float>();
 
@@ -134,12 +133,6 @@ public class LegacyPDFStreamEngine extends PDFStreamEngine {
         this.pageRotation = page.getRotation();
         this.pageSize = page.getCropBox();
 
-        if (pageSize.getLowerLeftX() == 0 && pageSize.getLowerLeftY() == 0) {
-            translateMatrix = null;
-        } else {
-            // translation matrix for cropbox
-            translateMatrix = Matrix.getTranslateInstance(-pageSize.getLowerLeftX(), -pageSize.getLowerLeftY());
-        }
         super.processPage(page);
     }
 
@@ -264,22 +257,13 @@ public class LegacyPDFStreamEngine extends PDFStreamEngine {
                 return;
             }
         }
-        // adjust for cropbox if needed
-        Matrix translatedTextRenderingMatrix;
-        if (translateMatrix == null) {
-            translatedTextRenderingMatrix = textRenderingMatrix;
-        } else {
-            translatedTextRenderingMatrix = Matrix.concatenate(translateMatrix, textRenderingMatrix);
-            nextX -= pageSize.getLowerLeftX();
-            nextY -= pageSize.getLowerLeftY();
-        }
 
         // This is a hack for unicode letter with 2 chars e.g. RA see unicodeProblem.pdf
         if (unicodeMapping.length() == 2) {
             processTextPosition(new TextPosition(pageRotation,
                                                  pageSize.getWidth(),
                                                  pageSize.getHeight(),
-                                                 translatedTextRenderingMatrix,
+                                                 textRenderingMatrix,
                                                  nextX,
                                                  nextY,
                                                  Math.abs(dyDisplay),
@@ -293,7 +277,7 @@ public class LegacyPDFStreamEngine extends PDFStreamEngine {
             processTextPosition(new TextPosition(pageRotation,
                                                  pageSize.getWidth(),
                                                  pageSize.getHeight(),
-                                                 translatedTextRenderingMatrix,
+                                                 textRenderingMatrix,
                                                  nextX,
                                                  nextY,
                                                  Math.abs(dyDisplay),
@@ -309,7 +293,7 @@ public class LegacyPDFStreamEngine extends PDFStreamEngine {
             processTextPosition(new TextPosition(pageRotation,
                                                  pageSize.getWidth(),
                                                  pageSize.getHeight(),
-                                                 translatedTextRenderingMatrix,
+                                                 textRenderingMatrix,
                                                  nextX,
                                                  nextY,
                                                  Math.abs(dyDisplay),

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/utils/ParsingConstants.java

@@ -1,7 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.utils;
-
-public class ParsingConstants {
-
-    public final static float NEW_LINE_TEXT_HEIGHT_PERCENTAGE = 0.6f;
-
-}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/utils/TextPositionSequenceComparator.java

@@ -51,47 +51,13 @@ public class TextPositionSequenceComparator implements Comparator<TextPositionSe
 
         double yDifference = Math.abs(pos1YBottom - pos2YBottom);
 
-        // Adjust for text rotation
-        switch (pos1.getRotation()) {
-            case 0:
-                // 0 degrees (horizontal, top to bottom and left to right): Sort primarily by y-coordinates from top to bottom (pos1YBottom < pos2YBottom).
-                if (yDifference < .1 || (pos2YBottom >= pos1YTop && pos2YBottom <= pos1YBottom) || (pos1YBottom >= pos2YTop && pos1YBottom <= pos2YBottom)) {
-                    return Double.compare(x1, x2);
-                } else if (pos1YBottom < pos2YBottom) {
-                    return -1;
-                } else {
-                    return 1;
-                }
-            case 90:
-                // 90 degrees (vertical, right to left): Sort by x-coordinates first (x1 > x2), then by y-coordinates from top to bottom (pos1YBottom < pos2YBottom).
-                if (x1 > x2) {
-                    return -1;
-                } else if (x1 < x2) {
-                    return 1;
-                } else {
-                    return Double.compare(pos1YBottom, pos2YBottom);
-                }
-            case 180:
-                // 180 degrees (horizontal, bottom to top and right to left): Sort primarily by y-coordinates from bottom to top (pos1YBottom > pos2YBottom).
-                if (yDifference < .1 || (pos2YBottom >= pos1YTop && pos2YBottom <= pos1YBottom) || (pos1YBottom >= pos2YTop && pos1YBottom <= pos2YBottom)) {
-                    return Double.compare(x2, x1);
-
-                } else if (pos1YBottom > pos2YBottom) {
-                    return -1;
-                } else {
-                    return 1;
-                }
-            case 270:
-                // 270 degrees (vertical, left to right): Sort by x-coordinates in reverse (x2 > x1), then by y-coordinates from bottom to top (pos2YBottom > pos1YBottom).
-                if (x2 > x1) {
-                    return -1;
-                } else if (x2 < x1) {
-                    return 1;
-                } else {
-                    return Double.compare(pos2YBottom, pos1YBottom);
-                }
-            default:
-                throw new RuntimeException("Rotation not supported. Only 0/90/180/270 degree rotation is supported.");
+        // we will do a simple tolerance comparison
+        if (yDifference < .1 || pos2YBottom >= pos1YTop && pos2YBottom <= pos1YBottom || pos1YBottom >= pos2YTop && pos1YBottom <= pos2YBottom) {
+            return Double.compare(x1, x2);
+        } else if (pos1YBottom < pos2YBottom) {
+            return -1;
+        } else {
+            return 1;
         }
     }
 

layoutparser-service/layoutparser-service-server/build.gradle.kts

@@ -32,7 +32,6 @@ dependencies {
     implementation("com.iqser.red.commons:storage-commons:2.45.0")
     implementation("com.knecon.fforesight:tenant-commons:0.21.0")
     implementation("com.knecon.fforesight:tracing-commons:0.5.0")
-    implementation("com.knecon.fforesight:lifecycle-commons:0.6.0")
     implementation("org.springframework.boot:spring-boot-starter-actuator:${springBootStarterVersion}")
     implementation("org.springframework.boot:spring-boot-starter-amqp:${springBootStarterVersion}")
     implementation("com.amazonaws:aws-java-sdk-s3:1.12.536")

layoutparser-service/layoutparser-service-server/src/main/java/com/knecon/fforesight/service/layoutparser/server/Application.java

@@ -6,22 +6,19 @@ import org.springframework.boot.autoconfigure.ImportAutoConfiguration;
 import org.springframework.boot.autoconfigure.SpringBootApplication;
 import org.springframework.boot.autoconfigure.security.servlet.SecurityAutoConfiguration;
 import org.springframework.context.annotation.Bean;
-import org.springframework.context.annotation.EnableAspectJAutoProxy;
 import org.springframework.context.annotation.Import;
 
 import com.amazonaws.services.s3.model.metrics.MetricsConfiguration;
 import com.iqser.red.storage.commons.StorageAutoConfiguration;
-import com.knecon.fforesight.lifecyclecommons.LifecycleAutoconfiguration;
 import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingServiceProcessorConfiguration;
 import com.knecon.fforesight.tenantcommons.MultiTenancyAutoConfiguration;
 
 import io.micrometer.observation.ObservationRegistry;
 import io.micrometer.observation.aop.ObservedAspect;
 
-@ImportAutoConfiguration({MultiTenancyAutoConfiguration.class, LifecycleAutoconfiguration.class})
+@ImportAutoConfiguration({MultiTenancyAutoConfiguration.class})
 @Import({MetricsConfiguration.class, StorageAutoConfiguration.class, LayoutParsingServiceProcessorConfiguration.class})
 @SpringBootApplication(exclude = {SecurityAutoConfiguration.class, ManagementWebSecurityAutoConfiguration.class})
-@EnableAspectJAutoProxy
 public class Application {
 
     public static void main(String[] args) {

layoutparser-service/layoutparser-service-server/src/main/resources/application.yml

@@ -15,9 +15,6 @@ project.version: 1.0-SNAPSHOT
 server:
   port: 8080
 
-lifecycle:
-  base-package: com.knecon.fforesight.service.layoutparser
-
 spring:
   application:
     name: layoutparser-service

layoutparser-service/viewer-doc-processor/src/main/java/com/knecon/fforesight/service/viewerdoc/service/ViewerDocumentService.java

@@ -1,7 +1,6 @@
 package com.knecon.fforesight.service.viewerdoc.service;
 
 import java.awt.geom.AffineTransform;
-import java.awt.geom.Line2D;
 import java.awt.geom.Rectangle2D;
 import java.io.File;
 import java.io.FileOutputStream;
@@ -85,7 +84,6 @@ public class ViewerDocumentService {
             pdPage.setContents(ContentStreamUtility.removeLayerFromContentStreams(allLayers, classifiers));
 
             AffineTransform textDeRotationMatrix = getTextDeRotationTransform(pdPage);
-            AffineTransform pageTransformationMatrix = getPageTransformationMatrix(pdPage);
 
             if (!ContentStreamClassifier.areAllContentStreamsEscaped(classifiers)) {
                 // We need to save the graphics state before, such that our appended content cannot be affected by previous content streams with side effects,
@@ -108,11 +106,7 @@ public class ViewerDocumentService {
 
                     contentStream.saveGraphicsState();
 
-                    drawVisualizationsToContentStream(pdDocument,
-                                                      visualization.getVisualizationsOnPages().get(pageNumber),
-                                                      contentStream,
-                                                      textDeRotationMatrix,
-                                                      pageTransformationMatrix);
+                    drawVisualizationsToContentStream(pdDocument, visualization.getVisualizationsOnPages().get(pageNumber), contentStream, textDeRotationMatrix);
 
                     contentStream.restoreGraphicsState();
 
@@ -139,12 +133,6 @@ public class ViewerDocumentService {
     }
 
 
-    private AffineTransform getPageTransformationMatrix(PDPage pdPage) {
-
-        return new AffineTransform(1, 0, 0, 1, pdPage.getCropBox().getLowerLeftX(), pdPage.getCropBox().getLowerLeftY());
-    }
-
-
     private static Map<ContentStreams.Identifier, PDOptionalContentGroup> addLayersToDocument(List<Visualizations> visualizations, PDDocument pdDocument) {
 
         Map<ContentStreams.Identifier, PDOptionalContentGroup> optionalContentGroupMap = new HashMap<>();
@@ -159,8 +147,7 @@ public class ViewerDocumentService {
     private static void drawVisualizationsToContentStream(PDDocument pdDocument,
                                                           VisualizationsOnPage visualizationsOnPage,
                                                           PDPageContentStream contentStream,
-                                                          AffineTransform textDeRotationMatrix,
-                                                          AffineTransform pageTransformationMatrix) throws IOException {
+                                                          AffineTransform textDeRotationMatrix) throws IOException {
 
         if (visualizationsOnPage.isMakePathsInvisible()) {
             contentStream.addRect(0, 0, 1, 1);
@@ -168,18 +155,17 @@ public class ViewerDocumentService {
         }
 
         for (ColoredLine coloredLine : visualizationsOnPage.getColoredLines()) {
-            Line2D line = transformLine(coloredLine.line(), pageTransformationMatrix);
             contentStream.setLineWidth(coloredLine.lineWidth());
             contentStream.setStrokingColor(coloredLine.color());
-            contentStream.moveTo((float) line.getX1(), (float) line.getY1());
-            contentStream.lineTo((float) line.getX2(), (float) line.getY2());
+            contentStream.moveTo((float) coloredLine.line().getX1(), (float) coloredLine.line().getY1());
+            contentStream.lineTo((float) coloredLine.line().getX2(), (float) coloredLine.line().getY2());
             contentStream.stroke();
         }
 
         for (ColoredRectangle coloredRectangle : visualizationsOnPage.getColoredRectangles()) {
             contentStream.setLineWidth(coloredRectangle.lineWidth());
             contentStream.setStrokingColor(coloredRectangle.color());
-            Rectangle2D r = transformRect(coloredRectangle.rectangle2D(), pageTransformationMatrix);
+            Rectangle2D r = coloredRectangle.rectangle2D();
             contentStream.addRect((float) r.getX(), (float) r.getY(), (float) r.getWidth(), (float) r.getHeight());
             contentStream.stroke();
         }
@@ -189,7 +175,7 @@ public class ViewerDocumentService {
             PDExtendedGraphicsState graphicsState = new PDExtendedGraphicsState();
             graphicsState.setNonStrokingAlphaConstant(filledRectangle.alpha());
             contentStream.setGraphicsStateParameters(graphicsState);
-            Rectangle2D r = transformRect(filledRectangle.rectangle2D(), pageTransformationMatrix);
+            Rectangle2D r = filledRectangle.rectangle2D();
             contentStream.addRect((float) r.getX(), (float) r.getY(), (float) r.getWidth(), (float) r.getHeight());
             contentStream.fill();
         }
@@ -204,7 +190,7 @@ public class ViewerDocumentService {
             } else {
                 contentStream.setRenderingMode(RenderingMode.FILL);
             }
-            Matrix textMatrix = getTextMatrix(placedText, textDeRotationMatrix, pageTransformationMatrix);
+            Matrix textMatrix = getTextMatrix(placedText, textDeRotationMatrix);
             contentStream.setTextMatrix(textMatrix);
             contentStream.showText(placedText.text());
             contentStream.endText();
@@ -237,7 +223,7 @@ public class ViewerDocumentService {
     }
 
 
-    private static Matrix getTextMatrix(PlacedText placedText, AffineTransform textDeRotationMatrix, AffineTransform pageTransformationMatrix) {
+    private static Matrix getTextMatrix(PlacedText placedText, AffineTransform textDeRotationMatrix) {
 
         Matrix textMatrix;
         if (placedText.textMatrix().isEmpty()) {
@@ -250,7 +236,7 @@ public class ViewerDocumentService {
         } else {
             textMatrix = placedText.textMatrix().get();
         }
-        return new Matrix(pageTransformationMatrix).multiply(textMatrix);
+        return textMatrix;
     }
 
 
@@ -327,16 +313,4 @@ public class ViewerDocumentService {
         });
     }
 
-
-    private static Line2D transformLine(Line2D line, AffineTransform pageTransformation) {
-
-        return new Line2D.Double(pageTransformation.transform(line.getP1(), null), pageTransformation.transform(line.getP2(), null));
-    }
-
-
-    private static Rectangle2D transformRect(Rectangle2D r, AffineTransform pageTransformation) {
-
-        return pageTransformation.createTransformedShape(r).getBounds2D();
-    }
-
 }