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

RED-9129: Track pagestatistics also for landscape pages in documine

Browse Source
This commit is contained in:
Dominique Eifländer 2024-05-08 10:50:36 +02:00
parent ae6bad830e
commit 71025f7f16
1 changed files with 40 additions and 36 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

18
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/LayoutParsingPipeline.java
View File

@ -83,13 +83,17 @@ public class LayoutParsingPipeline {
try (PDDocument originDocument = layoutParsingStorageService.getOriginFile(layoutParsingRequest.originFileStorageId())) { try (PDDocument originDocument = layoutParsingStorageService.getOriginFile(layoutParsingRequest.originFileStorageId())) {
ImageServiceResponse imageServiceResponse = new ImageServiceResponse(); ImageServiceResponse imageServiceResponse = new ImageServiceResponse();
if (layoutParsingRequest.imagesFileStorageId().isPresent()) { if (layoutParsingRequest.imagesFileStorageId()
imageServiceResponse = layoutParsingStorageService.getImagesFile(layoutParsingRequest.imagesFileStorageId().get()); .isPresent()) {
imageServiceResponse = layoutParsingStorageService.getImagesFile(layoutParsingRequest.imagesFileStorageId()
.get());
} }
TableServiceResponse tableServiceResponse = new TableServiceResponse(); TableServiceResponse tableServiceResponse = new TableServiceResponse();
if (layoutParsingRequest.tablesFileStorageId().isPresent()) { if (layoutParsingRequest.tablesFileStorageId()
tableServiceResponse = layoutParsingStorageService.getTablesFile(layoutParsingRequest.tablesFileStorageId().get()); .isPresent()) {
tableServiceResponse = layoutParsingStorageService.getTablesFile(layoutParsingRequest.tablesFileStorageId()
.get());
} }
ClassificationDocument classificationDocument = parseLayout(layoutParsingRequest.layoutParsingType(), originDocument, imageServiceResponse, tableServiceResponse); ClassificationDocument classificationDocument = parseLayout(layoutParsingRequest.layoutParsingType(), originDocument, imageServiceResponse, tableServiceResponse);
@ -213,7 +217,7 @@ public class LayoutParsingPipeline {
tableExtractionService.extractTables(cleanRulings, classificationPage); tableExtractionService.extractTables(cleanRulings, classificationPage);
buildPageStatistics(classificationPage); buildPageStatistics(classificationPage);
increaseDocumentStatistics(classificationPage, classificationDocument); increaseDocumentStatistics(layoutParsingType, classificationPage, classificationDocument);
classificationPages.add(classificationPage); classificationPages.add(classificationPage);
} }
@ -242,9 +246,9 @@ public class LayoutParsingPipeline {
} }
private void increaseDocumentStatistics(ClassificationPage classificationPage, ClassificationDocument document) { private void increaseDocumentStatistics(LayoutParsingType layoutParsingType, ClassificationPage classificationPage, ClassificationDocument document) {
if (!classificationPage.isLandscape()) { if (layoutParsingType.equals(LayoutParsingType.DOCUMINE) || !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());