Marmelator/redaction-service
1
0
Fork 0
Code Issues Pull Requests 14 Actions Packages Projects Releases 622 Wiki Activity

Fixed duplicate Textblock in Tables

Browse Source
This commit is contained in:
deiflaender 2020-12-03 15:03:55 +01:00
parent 8619e51480
commit c90eee23c4
1 changed files with 9 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

19
redaction-service-v1/redaction-service-server-v1/src/main/java/com/iqser/red/service/redaction/v1/server/tableextraction/service/TableExtractionService.java
View File

@ -30,12 +30,14 @@ public class TableExtractionService {
List<Cell> cells = findCells(cleanRulings.getHorizontal(), cleanRulings.getVertical());
List<TextBlock> toBeRemoved = new ArrayList<>();
for (AbstractTextContainer abstractTextContainer : page.getTextBlocks()) {
TextBlock textBlock = (TextBlock) abstractTextContainer;
for (Cell cell : cells) {
if (cell.intersects(textBlock.getMinX(), textBlock.getMinY(), textBlock.getWidth(),
textBlock.getHeight())) {
if (cell.intersects(textBlock.getMinX(), textBlock.getMinY(), textBlock.getWidth(), textBlock.getHeight())) {
cell.addTextBlock(textBlock);
toBeRemoved.add(textBlock);
break;
}
}
@ -44,7 +46,6 @@ public class TableExtractionService {
cells = new ArrayList<>(new HashSet<>(cells));
Utils.sort(cells, Rectangle.ILL_DEFINED_ORDER);
List<Rectangle> spreadsheetAreas = findSpreadsheetsFromCells(cells).stream()
.filter(r -> r.getWidth() > 0f && r.getHeight() > 0f)
.collect(Collectors.toList());
@ -67,25 +68,23 @@ public class TableExtractionService {
Iterator<AbstractTextContainer> itty = page.getTextBlocks().iterator();
while (itty.hasNext()) {
AbstractTextContainer textBlock = itty.next();
if (table.contains(textBlock)) {
if (position == -1) {
position = page.getTextBlocks().indexOf(textBlock);
}
itty.remove();
if (table.contains(textBlock) && position == -1) {
position = page.getTextBlocks().indexOf(textBlock);
}
}
if (position != -1) {
page.getTextBlocks().add(position, table);
}
}
page.getTextBlocks().removeAll(toBeRemoved);
}
public List<Cell> findCells(List<Ruling> horizontalRulingLines, List<Ruling> verticalRulingLines) {
List<Cell> cellsFound = new ArrayList<>();
Map<Point2D, Ruling[]> intersectionPoints = Ruling.findIntersections(horizontalRulingLines,
verticalRulingLines);
Map<Point2D, Ruling[]> intersectionPoints = Ruling.findIntersections(horizontalRulingLines, verticalRulingLines);
List<Point2D> intersectionPointsList = new ArrayList<>(intersectionPoints.keySet());
intersectionPointsList.sort(POINT_COMPARATOR);