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

RED-7607 - Rotating pages leads to lost annotations (RM & DM)

Browse Source 
- use getXDirAdj instead of getX
- add fontSizeCounter for landscape pages also
This commit is contained in:
Corina Olariu 2023-10-04 14:13:38 +03:00
parent f2c0991987
commit 99ed331a1e
3 changed files with 7 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/LayoutParsingPipeline.java
View File

@ -249,9 +249,9 @@ public class LayoutParsingPipeline {
private void increaseDocumentStatistics(ClassificationPage classificationPage, ClassificationDocument document) { private void increaseDocumentStatistics(ClassificationPage classificationPage, ClassificationDocument document) {
if (!classificationPage.isLandscape()) { // if (!classificationPage.isLandscape()) {
document.getFontSizeCounter().addAll(classificationPage.getFontSizeCounter().getCountPerValue()); document.getFontSizeCounter().addAll(classificationPage.getFontSizeCounter().getCountPerValue());
} // }
document.getFontCounter().addAll(classificationPage.getFontCounter().getCountPerValue()); document.getFontCounter().addAll(classificationPage.getFontCounter().getCountPerValue());
document.getTextHeightCounter().addAll(classificationPage.getTextHeightCounter().getCountPerValue()); document.getTextHeightCounter().addAll(classificationPage.getTextHeightCounter().getCountPerValue());
document.getFontStyleCounter().addAll(classificationPage.getFontStyleCounter().getCountPerValue()); document.getFontStyleCounter().addAll(classificationPage.getFontStyleCounter().getCountPerValue());

2
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/parsing/PDFTextStripper.java
View File

@ -1711,7 +1711,7 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
int numberOfStrings = line.size(); int numberOfStrings = line.size();
for (int i = 0; i < numberOfStrings; i++) { for (int i = 0; i < numberOfStrings; i++) {
WordWithTextPositions word = line.get(i); WordWithTextPositions word = line.get(i);
word.getTextPositions().sort(Comparator.comparing(TextPosition::getX)); word.getTextPositions().sort(Comparator.comparing(TextPosition::getXDirAdj));
writeString(word.getText(), word.getTextPositions(), isParagraphEnd && i == numberOfStrings - 1); writeString(word.getText(), word.getTextPositions(), isParagraphEnd && i == numberOfStrings - 1);
if (i < numberOfStrings - 1) { if (i < numberOfStrings - 1) {
writeWordSeparator(); writeWordSeparator();

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

@ -19,9 +19,10 @@ public final class PositionUtils {
double threshold = textBlock.getMostPopularWordHeight() * 3; double threshold = textBlock.getMostPopularWordHeight() * 3;
if (textBlock.getPdfMinX() + threshold > btf.getTopLeft().getX() && textBlock.getPdfMaxX() - threshold < btf.getTopLeft() if (textBlock.getPdfMinX() + threshold > btf.getTopLeft().getX()
.getX() + btf.getWidth() && textBlock.getPdfMinY() + threshold > btf.getTopLeft().getY() && textBlock.getPdfMaxY() - threshold < btf.getTopLeft() && textBlock.getPdfMaxX() - threshold < btf.getTopLeft().getX() + btf.getWidth()
.getY() + btf.getHeight()) { && textBlock.getPdfMinY() + threshold > btf.getTopLeft().getY()
&& textBlock.getPdfMaxY() - threshold < btf.getTopLeft().getY() + btf.getHeight()) {
return true; return true;
} else { } else {
return false; return false;