From a9338262c5249215936185bc228d4ede199f7e12 Mon Sep 17 00:00:00 2001
From: maverickstuder <maverick.studer@knecon.com>
Date: Tue, 7 May 2024 15:51:54 +0200
Subject: [PATCH] RED-7074: Design Subsection section tree structure algorithm
 * fix for boundary error

---
 .../textblock/ConcatenatedTextBlock.java      |  3 +-
 .../model/outline/TOCEnrichmentService.java   | 38 ++++++++++---------
 2 files changed, 21 insertions(+), 20 deletions(-)

diff --git a/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/textblock/ConcatenatedTextBlock.java b/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/textblock/ConcatenatedTextBlock.java
index 7038dbf..d48170b 100644
--- a/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/textblock/ConcatenatedTextBlock.java
+++ b/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/textblock/ConcatenatedTextBlock.java
@@ -56,8 +56,7 @@ public class ConcatenatedTextBlock implements TextBlock {
             boundary.setStart(start);
             boundary.setEnd(end);
         } else if (boundary.end() != start) {
-            //throw new UnsupportedOperationException(format("Can only concat consecutive TextBlocks, trying to concat %s and %s", boundary, textBlock.getBoundary()));
-            return this;
+            throw new UnsupportedOperationException(format("Can only concat consecutive TextBlocks, trying to concat %s and %s", boundary, textBlock.getBoundary()));
         }
         this.atomicTextBlocks.addAll(textBlock.getAtomicTextBlocks());
         boundary.setEnd(end);
diff --git a/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/TOCEnrichmentService.java b/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/TOCEnrichmentService.java
index 95849b4..4e257cb 100644
--- a/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/TOCEnrichmentService.java
+++ b/layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/outline/TOCEnrichmentService.java
@@ -2,6 +2,7 @@ package com.knecon.fforesight.service.layoutparser.processor.model.outline;
 
 import java.util.ArrayList;
 import java.util.Collections;
+import java.util.Iterator;
 import java.util.List;
 
 import org.springframework.stereotype.Service;
@@ -26,6 +27,11 @@ public class TOCEnrichmentService {
     public void assignSectionBlocksAndImages(ClassificationDocument document) {
 
         TableOfContents toc = document.getTableOfContents();
+        Iterator<TableOfContentItem> iterator = toc.iterator();
+        TableOfContentItem currentTOCItem = null;
+        if(iterator.hasNext()) {
+            currentTOCItem = iterator.next();
+        }
         List<AbstractPageBlock> startBlocks = new ArrayList<>();
         List<ClassifiedImage> startImages = new ArrayList<>();
         TableOfContentItem currentSection = null;
@@ -64,28 +70,24 @@ public class TOCEnrichmentService {
                     }
                     previousTable = table;
                 }
-                boolean matched = false;
 
-                for (TableOfContentItem tocItem : toc) {
-                    if (current instanceof TextPageBlock && tocItem.getHeadline().getText().equals(current.getText())) {
-                        if (!foundFirstHeadline) {
-                            foundFirstHeadline = true;
-                        }
-                        currentSection = tocItem;
-                        //sectionsMap.get(tocItem).add(current);
-                        tocItem.getSectionBlocks().add(current);
-                        currentPageTOCItems.add(tocItem);
-                        matched = true;
-                        break;
+                if (current instanceof TextPageBlock && currentTOCItem != null && currentTOCItem.getHeadline().getText().equals(current.getText())) {
+                    if (!foundFirstHeadline) {
+                        foundFirstHeadline = true;
+                    }
+                    currentSection = currentTOCItem;
+                    currentTOCItem.getSectionBlocks().add(current);
+                    currentPageTOCItems.add(currentTOCItem);
+
+                    if(iterator.hasNext()) {
+                        currentTOCItem = iterator.next();
                     }
                 }
 
-                if (!matched) {
-                    if (!foundFirstHeadline) {
-                        startBlocks.add(current);
-                    } else {
-                        currentSection.getSectionBlocks().add(current);
-                    }
+                if (!foundFirstHeadline) {
+                    startBlocks.add(current);
+                } else {
+                    currentSection.getSectionBlocks().add(current);
                 }
             }