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

RED-7375: table extractor prototype

Browse Source 
added layer for table extractor results in ViewerDocumentService
This commit is contained in:
yhampe 2024-01-15 09:43:23 +01:00
parent 0d4800622d
commit b47b187c8a
2 changed files with 27 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

30
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/visualization/ViewerDocumentService.java
View File

@ -66,6 +66,7 @@ public class ViewerDocumentService {
// If we collect all COSDictionaries we changed and tell it explicitly to only add the changed ones by using saveIncremental it's very fast. // If we collect all COSDictionaries we changed and tell it explicitly to only add the changed ones by using saveIncremental it's very fast.
Set<COSDictionary> dictionariesToUpdate = new HashSet<>(); Set<COSDictionary> dictionariesToUpdate = new HashSet<>();
PDOptionalContentGroup layer = addLayerToDocument(pdDocument, dictionariesToUpdate, layerVisibilityDefaultValue); PDOptionalContentGroup layer = addLayerToDocument(pdDocument, dictionariesToUpdate, layerVisibilityDefaultValue);
PDOptionalContentGroup tableExtractorLayer = addLayerToDocument(pdDocument, dictionariesToUpdate, true);
PDFont font = new PDType1Font(Standard14Fonts.FontName.HELVETICA); PDFont font = new PDType1Font(Standard14Fonts.FontName.HELVETICA);
for (int pageNumber = 0; pageNumber < pdDocument.getNumberOfPages(); pageNumber++) { for (int pageNumber = 0; pageNumber < pdDocument.getNumberOfPages(); pageNumber++) {
@ -108,11 +109,6 @@ public class ViewerDocumentService {
contentStream.addRect((float) r.getX(), (float) r.getY(), (float) r.getWidth(), (float) r.getHeight()); contentStream.addRect((float) r.getX(), (float) r.getY(), (float) r.getWidth(), (float) r.getHeight());
contentStream.fill(); contentStream.fill();
} }
for(TableCells tableCells: extractedTableCells.get(pageNumber)) {
contentStream.setStrokingColor(Color.CYAN);
contentStream.addRect((float) tableCells.getX0(), (float) tableCells.getY0(), (float) tableCells.getWidth(), (float) tableCells.getHeight());
contentStream.stroke();
}
for (PlacedText placedText : visualizationsOnPage.getPlacedTexts()) { for (PlacedText placedText : visualizationsOnPage.getPlacedTexts()) {
contentStream.setFont(font, FONT_SIZE); contentStream.setFont(font, FONT_SIZE);
contentStream.beginText(); contentStream.beginText();
@ -129,6 +125,30 @@ public class ViewerDocumentService {
} }
contentStream.restoreGraphicsState(); contentStream.restoreGraphicsState();
contentStream.endMarkedContent(); contentStream.endMarkedContent();
contentStream.beginMarkedContent(COSName.OC, tableExtractorLayer);
contentStream.saveGraphicsState();
contentStream.setLineWidth(LINE_WIDTH);
for (TableCells tableCells : extractedTableCells.get(pageNumber)) {
contentStream.setStrokingColor(new Color(0xFF0000));
contentStream.addRect((float) tableCells.getX0(), (float) tableCells.getY0(), (float) tableCells.getWidth(), (float) tableCells.getHeight());
contentStream.stroke();
contentStream.setFont(font, FONT_SIZE);
contentStream.beginText();
Matrix textMatrix = new Matrix((float) textDeRotationMatrix.getScaleX(),
(float) textDeRotationMatrix.getShearX(),
(float) textDeRotationMatrix.getShearY(),
(float) textDeRotationMatrix.getScaleY(),
tableCells.getX0() ,
tableCells.getY0());
// textMatrix.translate(-((font.getStringWidth(tableCells.getLabel()) / 1000) * FONT_SIZE + (2 * LINE_WIDTH) + 4), -FONT_SIZE);
// contentStream.setTextMatrix(textMatrix);
// contentStream.showText(tableCells.getLabel());
contentStream.endText();
}
contentStream.restoreGraphicsState();
contentStream.endMarkedContent();
} }
dictionariesToUpdate.add(pdPage.getCOSObject()); dictionariesToUpdate.add(pdPage.getCOSObject());
dictionariesToUpdate.add(pdPage.getResources().getCOSObject()); dictionariesToUpdate.add(pdPage.getResources().getCOSObject());

4
publish-custom-image.sh
View File

@ -4,5 +4,5 @@ gradle assemble
buildNumber=${1:-1} buildNumber=${1:-1}
gradle bootBuildImage --cleanCache --publishImage -PbuildbootDockerHostNetwork=true -Pversion=$USER-$buildNumber gradle bootBuildImage --cleanCache --publishImage -Pversion=layout-parser-yannik-$buildNumber --stacktrace
echo "nexus.knecon.com:5001/red/${dir}-server-v1:$USER-$buildNumber" echo "nexus.knecon.com:5001/red/${dir}-server-v1:table-extractor-yannik-$buildNumber"