Marmelator/layout-parser
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 289 Wiki Activity

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

Browse Source 
RED-9760: Fixed missing newLines

See merge request fforesight/layout-parser!185
This commit is contained in:
Dominique Eifländer 2024-07-30 16:06:30 +02:00
commit ca69d3f5dc
3 changed files with 17 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/text/TextPageBlock.java
View File

@ -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++;
}
}

11
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/factory/SearchTextWithTextPositionFactory.java
View File

@ -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();

7
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/utils/ParsingConstants.java Normal file
View File

@ -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;
}