clari-section-chunk: change structure object from paragraph/headline to sections

* add sections
This commit is contained in:
Kilian Schuettler 2024-06-26 12:35:14 +02:00
parent 80e701a0fd
commit 6b63596516

View File

@ -16,6 +16,7 @@ import com.knecon.fforesight.service.layoutparser.internal.api.data.taas.TableDa
import com.knecon.fforesight.service.layoutparser.processor.model.graph.Boundary;
import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Page;
import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Section;
import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SemanticNode;
import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Table;
import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.TableCell;
@ -29,10 +30,7 @@ public class TaasDocumentDataMapper {
List<StructureObject> structureObjects = document.streamAllSubNodes()
.filter(node -> !node.getType().equals(NodeType.TABLE_CELL))
.filter(node -> !node.getType().equals(NodeType.SUPER_SECTION))
.filter(node -> !node.getType().equals(NodeType.HEADER))
.filter(node -> !node.getType().equals(NodeType.FOOTER))
.filter(node -> !node.getType().equals(NodeType.HEADLINE))
.filter(node -> !node.getType().equals(NodeType.PARAGRAPH))
.filter(node -> !(node.getType().equals(NodeType.SECTION) && ((Section) node).hasTables()))
.map(node -> {
if (node.getType().equals(NodeType.TABLE)) {
return TaasDocumentDataMapper.fromTableWithTableData((Table) node, structureObjectNumber.getAndIncrement());