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

RED-7074: Design Subsection section tree structure algorithm

Browse Source 
* fix for boundary error
This commit is contained in:
maverickstuder 2024-05-07 15:51:54 +02:00
parent d2dc369df3
commit a9338262c5
2 changed files with 21 additions and 20 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

3
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/textblock/ConcatenatedTextBlock.java
View File

@ -56,8 +56,7 @@ public class ConcatenatedTextBlock implements TextBlock {
boundary.setStart(start); boundary.setStart(start);
boundary.setEnd(end); boundary.setEnd(end);
} else if (boundary.end() != start) { } else if (boundary.end() != start) {
//throw new UnsupportedOperationException(format("Can only concat consecutive TextBlocks, trying to concat %s and %s", boundary, textBlock.getBoundary())); throw new UnsupportedOperationException(format("Can only concat consecutive TextBlocks, trying to concat %s and %s", boundary, textBlock.getBoundary()));
return this;
} }
this.atomicTextBlocks.addAll(textBlock.getAtomicTextBlocks()); this.atomicTextBlocks.addAll(textBlock.getAtomicTextBlocks());
boundary.setEnd(end); boundary.setEnd(end);

24
layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/TOCEnrichmentService.java
View File

@ -2,6 +2,7 @@ package com.knecon.fforesight.service.layoutparser.processor.model.outline;
import java.util.ArrayList; import java.util.ArrayList;
import java.util.Collections; import java.util.Collections;
import java.util.Iterator;
import java.util.List; import java.util.List;
import org.springframework.stereotype.Service; import org.springframework.stereotype.Service;
@ -26,6 +27,11 @@ public class TOCEnrichmentService {
public void assignSectionBlocksAndImages(ClassificationDocument document) { public void assignSectionBlocksAndImages(ClassificationDocument document) {
TableOfContents toc = document.getTableOfContents(); TableOfContents toc = document.getTableOfContents();
Iterator<TableOfContentItem> iterator = toc.iterator();
TableOfContentItem currentTOCItem = null;
if(iterator.hasNext()) {
currentTOCItem = iterator.next();
}
List<AbstractPageBlock> startBlocks = new ArrayList<>(); List<AbstractPageBlock> startBlocks = new ArrayList<>();
List<ClassifiedImage> startImages = new ArrayList<>(); List<ClassifiedImage> startImages = new ArrayList<>();
TableOfContentItem currentSection = null; TableOfContentItem currentSection = null;
@ -64,30 +70,26 @@ public class TOCEnrichmentService {
} }
previousTable = table; previousTable = table;
} }
boolean matched = false;
for (TableOfContentItem tocItem : toc) { if (current instanceof TextPageBlock && currentTOCItem != null && currentTOCItem.getHeadline().getText().equals(current.getText())) {
if (current instanceof TextPageBlock && tocItem.getHeadline().getText().equals(current.getText())) {
if (!foundFirstHeadline) { if (!foundFirstHeadline) {
foundFirstHeadline = true; foundFirstHeadline = true;
} }
currentSection = tocItem; currentSection = currentTOCItem;
//sectionsMap.get(tocItem).add(current); currentTOCItem.getSectionBlocks().add(current);
tocItem.getSectionBlocks().add(current); currentPageTOCItems.add(currentTOCItem);
currentPageTOCItems.add(tocItem);
matched = true; if(iterator.hasNext()) {
break; currentTOCItem = iterator.next();
} }
} }
if (!matched) {
if (!foundFirstHeadline) { if (!foundFirstHeadline) {
startBlocks.add(current); startBlocks.add(current);
} else { } else {
currentSection.getSectionBlocks().add(current); currentSection.getSectionBlocks().add(current);
} }
} }
}
if (!currentPageTOCItems.isEmpty()) { if (!currentPageTOCItems.isEmpty()) {
lastFoundTOCItems = currentPageTOCItems; lastFoundTOCItems = currentPageTOCItems;