Merge branch 'RED-9760-4.1' into 'release/0.142.x'

RED-9760: Fixed missing newLines

See merge request fforesight/layout-parser!185

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

@@ -1,5 +1,7 @@
 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;
@@ -208,7 +210,7 @@ public class TextPageBlock extends AbstractPageBlock {
         TextPositionSequence previous = null;
         for (TextPositionSequence word : sequences) {
             if (previous != null) {
-                if (Math.abs(previous.getMaxYDirAdj() - word.getMaxYDirAdj()) > word.getTextHeight()) {
+                if (Math.abs(previous.getMaxYDirAdj() - word.getMaxYDirAdj()) > word.getTextHeight() * NEW_LINE_TEXT_HEIGHT_PERCENTAGE) {
                     sb.append('\n');
                 } else {
                     sb.append(' ');
@@ -228,7 +230,7 @@ public class TextPageBlock extends AbstractPageBlock {
         TextPositionSequence previous = null;
         for (TextPositionSequence word : sequences) {
             if (previous != null) {
-                if (word.getMaxYDirAdj() - previous.getMaxYDirAdj() > word.getTextHeight()) {
+                if (word.getMaxYDirAdj() - previous.getMaxYDirAdj() > word.getTextHeight() * NEW_LINE_TEXT_HEIGHT_PERCENTAGE) {
                     numberOfLines++;
                 }
             }

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

@@ -1,5 +1,7 @@
 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;
@@ -67,7 +69,6 @@ public class SearchTextWithTextPositionFactory {
             ++context.stringIdx;
         }
 
-
         List<Rectangle2D> positions = sequences.stream()
                 .map(TextPositionSequence::getTextPositions)
                 .flatMap(Collection::stream)
@@ -161,7 +162,7 @@ public class SearchTextWithTextPositionFactory {
         }
 
         double deltaY = Math.abs(currentPosition.getYDirAdj() - previousPosition.getYDirAdj());
-        return deltaY >= currentPosition.getHeightDir();
+        return deltaY >= currentPosition.getHeightDir() * NEW_LINE_TEXT_HEIGHT_PERCENTAGE;
     }
 
 
@@ -191,9 +192,9 @@ 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/utils/ParsingConstants.java

@@ -0,0 +1,7 @@
+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

@@ -51,13 +51,47 @@ public class TextPositionSequenceComparator implements Comparator<TextPositionSe
 
         double yDifference = Math.abs(pos1YBottom - pos2YBottom);
 
-        // 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;
+        // 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.");
         }
     }