Merge branch 'RED-1137' into 'main'

RED-1137: Do not observe actuator endpoints

See merge request fforesight/layout-parser!92

.gitlab-ci.yml

@@ -10,14 +10,12 @@ deploy:
   script:
     - echo "Building with gradle version ${BUILDVERSION}"
     - gradle -Pversion=${BUILDVERSION} publish
-    - gradle bootBuildImage --cleanCache --publishImage -PbuildbootDockerHostNetwork=true -Pversion=${BUILDVERSION}
+    - gradle bootBuildImage --publishImage -PbuildbootDockerHostNetwork=true -Pversion=${BUILDVERSION}
     - echo "BUILDVERSION=$BUILDVERSION" >> version.env
   artifacts:
     reports:
       dotenv: version.env
   rules:
-    - if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH
+    - if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH 
     - if: $CI_COMMIT_BRANCH =~ /^release/
     - if: $CI_COMMIT_TAG
-pmd:
-  allow_failure: true

buildSrc/src/main/kotlin/com.knecon.fforesight.java-conventions.gradle.kts

@@ -24,6 +24,8 @@ tasks.named<Test>("test") {
     reports {
         junitXml.outputLocation.set(layout.buildDirectory.dir("reports/junit"))
     }
+    minHeapSize = "512m"
+    maxHeapSize = "2048m"
 }
 
 tasks.test {

layoutparser-service/layoutparser-service-internal-api/build.gradle.kts

@@ -1,6 +1,6 @@
 plugins {
     id("com.knecon.fforesight.java-conventions")
-    id("io.freefair.lombok") version "8.6"
+    id("io.freefair.lombok") version "8.2.2"
 }
 
 description = "layoutparser-service-internal-api"

layoutparser-service/layoutparser-service-processor/build.gradle.kts

@@ -1,6 +1,6 @@
 plugins {
     id("com.knecon.fforesight.java-conventions")
-    id("io.freefair.lombok") version "8.6"
+    id("io.freefair.lombok") version "8.2.2"
 }
 
 description = "layoutparser-service-processor"
@@ -15,8 +15,8 @@ dependencies {
         exclude("org.springframework.boot", "spring-boot-starter-security")
         exclude("org.springframework.boot", "spring-boot-starter-validation")
     }
-    implementation("com.knecon.fforesight:tenant-commons:0.10.0")
-    implementation("com.iqser.red.commons:storage-commons:2.40.0")
+    implementation("com.knecon.fforesight:tenant-commons:0.19.0")
+    implementation("com.iqser.red.commons:storage-commons:2.45.0")
 
     implementation("org.apache.pdfbox:pdfbox:${pdfBoxVersion}")
     implementation("org.apache.pdfbox:pdfbox-tools:${pdfBoxVersion}")

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

@@ -3,13 +3,16 @@ package com.knecon.fforesight.service.layoutparser.processor;
 import static java.lang.String.format;
 
 import java.awt.geom.Rectangle2D;
-import java.io.ByteArrayOutputStream;
+import java.io.File;
 import java.io.IOException;
+import java.nio.file.Files;
 import java.util.ArrayList;
 import java.util.HashMap;
 import java.util.List;
 import java.util.Map;
+import java.util.concurrent.atomic.AtomicReference;
 
+import org.apache.pdfbox.Loader;
 import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.pdmodel.common.PDRectangle;
@@ -51,128 +54,169 @@ import com.knecon.fforesight.service.layoutparser.processor.services.parsing.PDF
 import com.knecon.fforesight.service.layoutparser.processor.services.visualization.ViewerDocumentService;
 import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 
+import io.micrometer.observation.Observation;
+import io.micrometer.observation.ObservationRegistry;
+import io.micrometer.observation.annotation.Observed;
+import lombok.AccessLevel;
 import lombok.RequiredArgsConstructor;
 import lombok.SneakyThrows;
+import lombok.experimental.FieldDefaults;
 import lombok.extern.slf4j.Slf4j;
 
+@SuppressWarnings("PMD.CloseResource")
 @Slf4j
 @Service
 @RequiredArgsConstructor
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
 public class LayoutParsingPipeline {
 
-    private final ImageServiceResponseAdapter imageServiceResponseAdapter;
-    private final CvTableParsingAdapter cvTableParsingAdapter;
-    private final LayoutParsingStorageService layoutParsingStorageService;
-    private final SectionsBuilderService sectionsBuilderService;
-    private final TaasClassificationService taasClassificationService;
-    private final RedactManagerClassificationService redactManagerClassificationService;
-    private final DocuMineClassificationService docuMineClassificationService;
-    private final SimplifiedSectionTextService simplifiedSectionTextService;
-    private final BodyTextFrameService bodyTextFrameService;
-    private final RulingCleaningService rulingCleaningService;
-    private final TableExtractionService tableExtractionService;
-    private final TaasBlockificationService taasBlockificationService;
-    private final DocuMineBlockificationService docuMineBlockificationService;
-    private final RedactManagerBlockificationService redactManagerBlockificationService;
-    private final ViewerDocumentService viewerDocumentService;
+    ImageServiceResponseAdapter imageServiceResponseAdapter;
+    CvTableParsingAdapter cvTableParsingAdapter;
+    LayoutParsingStorageService layoutParsingStorageService;
+    SectionsBuilderService sectionsBuilderService;
+    TaasClassificationService taasClassificationService;
+    RedactManagerClassificationService redactManagerClassificationService;
+    DocuMineClassificationService docuMineClassificationService;
+    SimplifiedSectionTextService simplifiedSectionTextService;
+    BodyTextFrameService bodyTextFrameService;
+    RulingCleaningService rulingCleaningService;
+    TableExtractionService tableExtractionService;
+    TaasBlockificationService taasBlockificationService;
+    DocuMineBlockificationService docuMineBlockificationService;
+    RedactManagerBlockificationService redactManagerBlockificationService;
+    ViewerDocumentService viewerDocumentService;
+    ObservationRegistry observationRegistry;
 
 
     public LayoutParsingFinishedEvent parseLayoutAndSaveFilesToStorage(LayoutParsingRequest layoutParsingRequest) throws IOException {
 
         long start = System.currentTimeMillis();
+        log.info("Starting layout parsing for {}", layoutParsingRequest.identifier());
 
-        try (PDDocument originDocument = layoutParsingStorageService.getOriginFile(layoutParsingRequest.originFileStorageId())) {
-            ImageServiceResponse imageServiceResponse = new ImageServiceResponse();
-            if (layoutParsingRequest.imagesFileStorageId()
-                    .isPresent()) {
-                imageServiceResponse = layoutParsingStorageService.getImagesFile(layoutParsingRequest.imagesFileStorageId()
-                                                                                         .get());
-            }
+        File originFile = layoutParsingStorageService.getOriginFile(layoutParsingRequest.originFileStorageId());
+        File viewerDocumentFile = File.createTempFile("viewer_document", ".pdf");
 
-            TableServiceResponse tableServiceResponse = new TableServiceResponse();
-            if (layoutParsingRequest.tablesFileStorageId()
-                    .isPresent()) {
-                tableServiceResponse = layoutParsingStorageService.getTablesFile(layoutParsingRequest.tablesFileStorageId()
-                                                                                         .get());
-            }
-
-            ClassificationDocument classificationDocument = parseLayout(layoutParsingRequest.layoutParsingType(), originDocument, imageServiceResponse, tableServiceResponse);
-            Document documentGraph = DocumentGraphFactory.buildDocumentGraph(classificationDocument);
-
-            int numberOfPages = originDocument.getNumberOfPages();
-
-            layoutParsingStorageService.storeDocumentData(layoutParsingRequest, DocumentDataMapper.toDocumentData(documentGraph));
-            layoutParsingStorageService.storeSimplifiedText(layoutParsingRequest, simplifiedSectionTextService.toSimplifiedText(documentGraph));
-
-            try (var out = new ByteArrayOutputStream()) {
-                viewerDocumentService.createViewerDocument(originDocument, documentGraph, out, false);
-                layoutParsingStorageService.storeViewerDocument(layoutParsingRequest, out);
-            }
-
-            if (layoutParsingRequest.layoutParsingType().equals(LayoutParsingType.TAAS)) {
-                var researchDocumentData = TaasDocumentDataMapper.fromDocument(documentGraph);
-                layoutParsingStorageService.storeResearchDocumentData(layoutParsingRequest, researchDocumentData);
-            }
-
-            return LayoutParsingFinishedEvent.builder()
-                    .identifier(layoutParsingRequest.identifier())
-                    .numberOfPages(numberOfPages)
-                    .duration(System.currentTimeMillis() - start)
-                    .message(format("""
-                                            Layout parsing has finished in %.02f s.
-                                            identifiers: %s
-                                            %s
-                                            Files have been saved with Ids:
-                                            Structure: %s
-                                            Text: %s
-                                            Positions: %s
-                                            PageData: %s
-                                            Simplified Text: %s
-                                            Viewer Doc: %s""",
-                                    ((float) (System.currentTimeMillis() - start)) / 1000,
-                                    layoutParsingRequest.identifier(),
-                                    buildSemanticNodeCountMessage(numberOfPages, documentGraph.buildSemanticNodeCounts()),
-                                    layoutParsingRequest.structureFileStorageId(),
-                                    layoutParsingRequest.textBlockFileStorageId(),
-                                    layoutParsingRequest.positionBlockFileStorageId(),
-                                    layoutParsingRequest.pageFileStorageId(),
-                                    layoutParsingRequest.simplifiedTextStorageId(),
-                                    layoutParsingRequest.viewerDocumentStorageId()))
-                    .build();
+        ImageServiceResponse imageServiceResponse = new ImageServiceResponse();
+        if (layoutParsingRequest.imagesFileStorageId().isPresent()) {
+            imageServiceResponse = layoutParsingStorageService.getImagesFile(layoutParsingRequest.imagesFileStorageId().get());
         }
+
+        TableServiceResponse tableServiceResponse = new TableServiceResponse();
+        if (layoutParsingRequest.tablesFileStorageId().isPresent()) {
+            tableServiceResponse = layoutParsingStorageService.getTablesFile(layoutParsingRequest.tablesFileStorageId().get());
+        }
+
+        ClassificationDocument classificationDocument = parseLayout(layoutParsingRequest.layoutParsingType(),
+                originFile,
+                imageServiceResponse,
+                tableServiceResponse,
+                layoutParsingRequest.identifier().toString());
+        log.info("Building document graph for {}", layoutParsingRequest.identifier());
+
+        Document documentGraph = observeBuildDocumentGraph(classificationDocument);
+
+        log.info("Storing resulting files for {}", layoutParsingRequest.identifier());
+        layoutParsingStorageService.storeDocumentData(layoutParsingRequest, DocumentDataMapper.toDocumentData(documentGraph));
+        layoutParsingStorageService.storeSimplifiedText(layoutParsingRequest, simplifiedSectionTextService.toSimplifiedText(documentGraph));
+
+        log.info("Creating viewer document for {}", layoutParsingRequest.identifier());
+        viewerDocumentService.createViewerDocument(originFile, documentGraph, viewerDocumentFile, false);
+        layoutParsingStorageService.storeViewerDocument(layoutParsingRequest, viewerDocumentFile);
+        if (layoutParsingRequest.layoutParsingType().equals(LayoutParsingType.TAAS)) {
+            log.info("Building research document data for {}", layoutParsingRequest.identifier());
+            var researchDocumentData = TaasDocumentDataMapper.fromDocument(documentGraph);
+            layoutParsingStorageService.storeResearchDocumentData(layoutParsingRequest, researchDocumentData);
+        }
+
+        originFile.delete();
+        viewerDocumentFile.delete();
+
+        return LayoutParsingFinishedEvent.builder()
+                .identifier(layoutParsingRequest.identifier())
+                .numberOfPages(documentGraph.getNumberOfPages())
+                .duration(System.currentTimeMillis() - start)
+                .message(format("""
+                                Layout parsing has finished in %.02f s.
+                                identifiers: %s
+                                %s
+                                Files have been saved with Ids:
+                                Structure: %s
+                                Text: %s
+                                Positions: %s
+                                PageData: %s
+                                Simplified Text: %s
+                                Viewer Doc: %s""",
+                        ((float) (System.currentTimeMillis() - start)) / 1000,
+                        layoutParsingRequest.identifier(),
+                        buildSemanticNodeCountMessage(documentGraph.getNumberOfPages(), documentGraph.buildSemanticNodeCounts()),
+                        layoutParsingRequest.structureFileStorageId(),
+                        layoutParsingRequest.textBlockFileStorageId(),
+                        layoutParsingRequest.positionBlockFileStorageId(),
+                        layoutParsingRequest.pageFileStorageId(),
+                        layoutParsingRequest.simplifiedTextStorageId(),
+                        layoutParsingRequest.viewerDocumentStorageId()))
+                .build();
+
+    }
+
+
+    private Document observeBuildDocumentGraph(ClassificationDocument classificationDocument) {
+
+        AtomicReference<Document> documentReference = new AtomicReference<>();
+
+        Observation.createNotStarted("LayoutParsingPipeline", observationRegistry).contextualName("build-document-graph").observe(() -> {
+            documentReference.set(DocumentGraphFactory.buildDocumentGraph(classificationDocument));
+        });
+
+        return documentReference.get();
     }
 
 
     private String buildSemanticNodeCountMessage(int numberOfPages, Map<NodeType, Long> semanticNodeCounts) {
 
         return String.format("%d pages with %d sections, %d headlines, %d paragraphs, %d tables with %d cells, %d headers, and %d footers parsed",
-                             numberOfPages,
-                             semanticNodeCounts.get(NodeType.SECTION) == null ? 0 : semanticNodeCounts.get(NodeType.SECTION),
-                             semanticNodeCounts.get(NodeType.HEADLINE) == null ? 0 : semanticNodeCounts.get(NodeType.HEADLINE),
-                             semanticNodeCounts.get(NodeType.PARAGRAPH) == null ? 0 : semanticNodeCounts.get(NodeType.PARAGRAPH),
-                             semanticNodeCounts.get(NodeType.TABLE) == null ? 0 : semanticNodeCounts.get(NodeType.TABLE),
-                             semanticNodeCounts.get(NodeType.TABLE_CELL) == null ? 0 : semanticNodeCounts.get(NodeType.TABLE_CELL),
-                             semanticNodeCounts.get(NodeType.HEADER) == null ? 0 : semanticNodeCounts.get(NodeType.HEADER),
-                             semanticNodeCounts.get(NodeType.FOOTER) == null ? 0 : semanticNodeCounts.get(NodeType.FOOTER));
+                numberOfPages,
+                semanticNodeCounts.get(NodeType.SECTION) == null ? 0 : semanticNodeCounts.get(NodeType.SECTION),
+                semanticNodeCounts.get(NodeType.HEADLINE) == null ? 0 : semanticNodeCounts.get(NodeType.HEADLINE),
+                semanticNodeCounts.get(NodeType.PARAGRAPH) == null ? 0 : semanticNodeCounts.get(NodeType.PARAGRAPH),
+                semanticNodeCounts.get(NodeType.TABLE) == null ? 0 : semanticNodeCounts.get(NodeType.TABLE),
+                semanticNodeCounts.get(NodeType.TABLE_CELL) == null ? 0 : semanticNodeCounts.get(NodeType.TABLE_CELL),
+                semanticNodeCounts.get(NodeType.HEADER) == null ? 0 : semanticNodeCounts.get(NodeType.HEADER),
+                semanticNodeCounts.get(NodeType.FOOTER) == null ? 0 : semanticNodeCounts.get(NodeType.FOOTER));
     }
 
 
     @SneakyThrows
+    @Observed(name = "LayoutParsingPipeline", contextualName = "parse-layout")
     public ClassificationDocument parseLayout(LayoutParsingType layoutParsingType,
-                                              PDDocument originDocument,
+                                              File originFile,
                                               ImageServiceResponse imageServiceResponse,
-                                              TableServiceResponse tableServiceResponse) {
+                                              TableServiceResponse tableServiceResponse,
+                                              String identifier) {
 
+        PDDocument originDocument = openDocument(originFile);
+        addNumberOfPagesToTrace(originDocument.getNumberOfPages(), Files.size(originFile.toPath()));
         Map<Integer, List<TableCells>> pdfTableCells = cvTableParsingAdapter.buildCvParsedTablesPerPage(tableServiceResponse);
         Map<Integer, List<ClassifiedImage>> pdfImages = imageServiceResponseAdapter.buildClassifiedImagesPerPage(imageServiceResponse);
 
         ClassificationDocument classificationDocument = new ClassificationDocument();
         List<ClassificationPage> classificationPages = new ArrayList<>();
 
-        originDocument.setAllSecurityToBeRemoved(true);
         long pageCount = originDocument.getNumberOfPages();
 
         for (int pageNumber = 1; pageNumber <= pageCount; pageNumber++) {
+
+            if (pageNumber % 100 == 0) {
+                // re-open document every once in a while to save on RAM. This has no significant performance impact.
+                // This is due to PDFBox caching all images and some other stuff with Soft References. This dereferences them and forces the freeing of memory.
+                originDocument.close();
+                originDocument = openDocument(originFile);
+            }
+
+            if (pageNumber % 100 == 0 || pageNumber == pageCount || pageNumber == 1) {
+                log.info("Extracting text on Page {} for {}", pageNumber, identifier);
+            }
+
             classificationDocument.setPages(classificationPages);
             PDFLinesTextStripper stripper = new PDFLinesTextStripper();
             PDPage pdPage = originDocument.getPage(pageNumber - 1);
@@ -217,26 +261,47 @@ public class LayoutParsingPipeline {
             tableExtractionService.extractTables(cleanRulings, classificationPage);
 
             buildPageStatistics(classificationPage);
-            increaseDocumentStatistics(layoutParsingType, classificationPage, classificationDocument);
+            increaseDocumentStatistics(classificationPage, classificationDocument);
 
             classificationPages.add(classificationPage);
         }
 
-        bodyTextFrameService.setBodyTextFrames(classificationDocument, layoutParsingType);
+        originDocument.close();
 
+        log.info("Calculating BodyTextFrame for {}", identifier);
+        bodyTextFrameService.setBodyTextFrames(classificationDocument, layoutParsingType);
+        log.info("Classify TextBlocks for {}", identifier);
         switch (layoutParsingType) {
             case TAAS -> taasClassificationService.classifyDocument(classificationDocument);
             case DOCUMINE -> docuMineClassificationService.classifyDocument(classificationDocument);
             case REDACT_MANAGER -> redactManagerClassificationService.classifyDocument(classificationDocument);
         }
 
+        log.info("Building Sections for {}", identifier);
         sectionsBuilderService.buildSections(classificationDocument);
         sectionsBuilderService.addImagesToSections(classificationDocument);
-
         return classificationDocument;
     }
 
 
+    private void addNumberOfPagesToTrace(int numberOfPages, long size) {
+
+        if (observationRegistry.getCurrentObservation() != null) {
+            observationRegistry.getCurrentObservation().highCardinalityKeyValue("numberOfPages", String.valueOf(numberOfPages));
+            observationRegistry.getCurrentObservation().highCardinalityKeyValue("fileSize", String.valueOf(size));
+        }
+    }
+
+
+    @SneakyThrows
+    private PDDocument openDocument(File originFile) {
+
+        PDDocument document = Loader.loadPDF(originFile);
+        document.setAllSecurityToBeRemoved(true);
+        return document;
+    }
+
+
     private Map<String, List<Rectangle2D>> convertMarkedContents(List<PDMarkedContent> pdMarkedContents) {
 
         Map<String, List<Rectangle2D>> markedContentBboxes = new HashMap<>();
@@ -246,9 +311,9 @@ public class LayoutParsingPipeline {
     }
 
 
-    private void increaseDocumentStatistics(LayoutParsingType layoutParsingType, ClassificationPage classificationPage, ClassificationDocument document) {
+    private void increaseDocumentStatistics(ClassificationPage classificationPage, ClassificationDocument document) {
 
-        if (layoutParsingType.equals(LayoutParsingType.DOCUMINE) || !classificationPage.isLandscape()) {
+        if (!classificationPage.isLandscape()) {
             document.getFontSizeCounter().addAll(classificationPage.getFontSizeCounter().getCountPerValue());
         }
         document.getFontCounter().addAll(classificationPage.getFontCounter().getCountPerValue());

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/LayoutParsingStorageService.java

@@ -1,9 +1,7 @@
 package com.knecon.fforesight.service.layoutparser.processor;
 
-import java.io.ByteArrayInputStream;
-import java.io.ByteArrayOutputStream;
 import java.io.File;
-import java.io.FileOutputStream;
+import java.io.FileInputStream;
 import java.io.IOException;
 import java.io.InputStream;
 import java.nio.file.Files;
@@ -11,7 +9,6 @@ import java.nio.file.Path;
 import java.nio.file.Paths;
 import java.nio.file.StandardOpenOption;
 
-import org.apache.commons.io.IOUtils;
 import org.apache.pdfbox.Loader;
 import org.apache.pdfbox.pdmodel.PDDocument;
 import org.springframework.stereotype.Service;
@@ -26,6 +23,7 @@ import com.knecon.fforesight.service.layoutparser.processor.python_api.model.ima
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
 import com.knecon.fforesight.tenantcommons.TenantContext;
 
+import io.micrometer.observation.annotation.Observed;
 import lombok.RequiredArgsConstructor;
 import lombok.SneakyThrows;
 import lombok.extern.slf4j.Slf4j;
@@ -39,16 +37,18 @@ public class LayoutParsingStorageService {
     private final ObjectMapper objectMapper;
 
 
-    public PDDocument getOriginFile(String storageId) throws IOException {
+    public PDDocument getOriginDocument(String storageId) throws IOException {
 
-        try (var originDocumentInputStream = getObject(storageId)) {
-            File tempFile = createTempFile("document", ".pdf");
-            try (var tempFileOutputStream = new FileOutputStream(tempFile)) {
-                IOUtils.copy(originDocumentInputStream, tempFileOutputStream);
-                originDocumentInputStream.close();
-            }
-            return Loader.loadPDF(tempFile);
-        }
+        return Loader.loadPDF(getOriginFile(storageId));
+    }
+
+
+    @Observed(name = "LayoutParsingStorageService", contextualName = "get-origin-file")
+    public File getOriginFile(String storageId) throws IOException {
+
+        File tempFile = createTempFile("document", ".pdf");
+        storageService.downloadTo(TenantContext.getTenantId(), storageId, tempFile);
+        return tempFile;
     }
 
 
@@ -74,6 +74,7 @@ public class LayoutParsingStorageService {
     }
 
 
+    @Observed(name = "LayoutParsingStorageService", contextualName = "store-document-data")
     public void storeDocumentData(LayoutParsingRequest layoutParsingRequest, DocumentData documentData) {
 
         storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.structureFileStorageId(), documentData.getDocumentStructure());
@@ -83,7 +84,6 @@ public class LayoutParsingStorageService {
     }
 
 
-
     public void storeResearchDocumentData(LayoutParsingRequest layoutParsingRequest, ResearchDocumentData researchDocumentData) {
 
         storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.researchDocumentStorageId(), researchDocumentData);
@@ -115,6 +115,7 @@ public class LayoutParsingStorageService {
     }
 
 
+    @Observed(name = "LayoutParsingStorageService", contextualName = "store-simplified-text")
     public void storeSimplifiedText(LayoutParsingRequest layoutParsingRequest, SimplifiedText simplifiedText) {
 
         storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.simplifiedTextStorageId(), simplifiedText);
@@ -132,9 +133,10 @@ public class LayoutParsingStorageService {
 
 
     @SneakyThrows
-    public void storeViewerDocument(LayoutParsingRequest layoutParsingRequest, ByteArrayOutputStream out) {
+    @Observed(name = "LayoutParsingStorageService", contextualName = "store-viewer-document")
+    public void storeViewerDocument(LayoutParsingRequest layoutParsingRequest, File out) {
 
-        try (var in = new ByteArrayInputStream(out.toByteArray())) {
+        try (var in = new FileInputStream(out)) {
 
             storageService.storeObject(TenantContext.getTenantId(), layoutParsingRequest.viewerDocumentStorageId(), in);
         }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Document.java

@@ -10,7 +10,6 @@ import java.util.Set;
 import java.util.stream.Collectors;
 import java.util.stream.Stream;
 
-import com.amazonaws.services.kms.model.NotFoundException;
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.DocumentTree;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.entity.RedactionEntity;
@@ -84,7 +83,7 @@ public class Document implements GenericSemanticNode {
     @Override
     public Headline getHeadline() {
 
-        return streamAllSubNodesOfType(NodeType.HEADLINE).map(node -> (Headline) node).findFirst().orElseThrow(() -> new NotFoundException("No Headlines found in this document!"));
+        return streamAllSubNodesOfType(NodeType.HEADLINE).map(node -> (Headline) node).findFirst().orElse(Headline.builder().build());
     }
 
 
@@ -105,6 +104,7 @@ public class Document implements GenericSemanticNode {
         return streamAllSubNodes().collect(Collectors.groupingBy(SemanticNode::getType, Collectors.counting()));
     }
 
+
     @Override
     public String toString() {
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Footer.java

@@ -1,7 +1,9 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
+import java.awt.geom.Rectangle2D;
 import java.util.HashSet;
 import java.util.List;
+import java.util.Map;
 import java.util.Set;
 
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
@@ -34,6 +36,9 @@ public class Footer implements GenericSemanticNode {
     @EqualsAndHashCode.Exclude
     Set<RedactionEntity> entities = new HashSet<>();
 
+    @EqualsAndHashCode.Exclude
+    Map<Page, Rectangle2D> bBoxCache;
+
 
     @Override
     public NodeType getType() {
@@ -62,4 +67,14 @@ public class Footer implements GenericSemanticNode {
         return treeId + ": " + NodeType.FOOTER + ": " + leafTextBlock.buildSummary();
     }
 
+
+    @Override
+    public Map<Page, Rectangle2D> getBBox() {
+
+        if (bBoxCache == null) {
+            bBoxCache = GenericSemanticNode.super.getBBox();
+        }
+        return bBoxCache;
+    }
+
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Header.java

@@ -1,7 +1,9 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
+import java.awt.geom.Rectangle2D;
 import java.util.HashSet;
 import java.util.List;
+import java.util.Map;
 import java.util.Set;
 
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
@@ -34,6 +36,9 @@ public class Header implements GenericSemanticNode {
     @EqualsAndHashCode.Exclude
     Set<RedactionEntity> entities = new HashSet<>();
 
+    @EqualsAndHashCode.Exclude
+    Map<Page, Rectangle2D> bBoxCache;
+
 
     @Override
     public boolean isLeaf() {
@@ -62,4 +67,14 @@ public class Header implements GenericSemanticNode {
         return treeId + ": " + NodeType.HEADER + ": " + leafTextBlock.buildSummary();
     }
 
+
+    @Override
+    public Map<Page, Rectangle2D> getBBox() {
+
+        if (bBoxCache == null) {
+            bBoxCache = GenericSemanticNode.super.getBBox();
+        }
+        return bBoxCache;
+    }
+
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Headline.java

@@ -1,7 +1,9 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
+import java.awt.geom.Rectangle2D;
 import java.util.HashSet;
 import java.util.List;
+import java.util.Map;
 import java.util.Set;
 
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
@@ -34,6 +36,9 @@ public class Headline implements GenericSemanticNode {
     @EqualsAndHashCode.Exclude
     Set<RedactionEntity> entities = new HashSet<>();
 
+    @EqualsAndHashCode.Exclude
+    Map<Page, Rectangle2D> bBoxCache;
+
 
     @Override
     public NodeType getType() {
@@ -69,4 +74,14 @@ public class Headline implements GenericSemanticNode {
         return this;
     }
 
+
+    @Override
+    public Map<Page, Rectangle2D> getBBox() {
+
+        if (bBoxCache == null) {
+            bBoxCache = GenericSemanticNode.super.getBBox();
+        }
+        return bBoxCache;
+    }
+
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Paragraph.java

@@ -1,7 +1,9 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
+import java.awt.geom.Rectangle2D;
 import java.util.HashSet;
 import java.util.List;
+import java.util.Map;
 import java.util.Set;
 
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
@@ -32,6 +34,9 @@ public class Paragraph implements GenericSemanticNode {
     @EqualsAndHashCode.Exclude
     Set<RedactionEntity> entities = new HashSet<>();
 
+    @EqualsAndHashCode.Exclude
+    Map<Page, Rectangle2D> bBoxCache;
+
 
     @Override
     public NodeType getType() {
@@ -60,4 +65,14 @@ public class Paragraph implements GenericSemanticNode {
         return treeId + ": " + NodeType.PARAGRAPH + ": " + leafTextBlock.buildSummary();
     }
 
+
+    @Override
+    public Map<Page, Rectangle2D> getBBox() {
+
+        if (bBoxCache == null) {
+            bBoxCache = GenericSemanticNode.super.getBBox();
+        }
+        return bBoxCache;
+    }
+
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Section.java

@@ -1,7 +1,9 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
+import java.awt.geom.Rectangle2D;
 import java.util.HashSet;
 import java.util.List;
+import java.util.Map;
 import java.util.Set;
 
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
@@ -35,6 +37,9 @@ public class Section implements GenericSemanticNode {
     @EqualsAndHashCode.Exclude
     Set<RedactionEntity> entities = new HashSet<>();
 
+    @EqualsAndHashCode.Exclude
+    Map<Page, Rectangle2D> bBoxCache;
+
 
     @Override
     public NodeType getType() {
@@ -74,4 +79,14 @@ public class Section implements GenericSemanticNode {
                 .orElseGet(() -> getParent().getHeadline());
     }
 
+
+    @Override
+    public Map<Page, Rectangle2D> getBBox() {
+
+        if (bBoxCache == null) {
+            bBoxCache = GenericSemanticNode.super.getBBox();
+        }
+        return bBoxCache;
+    }
+
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/SemanticNode.java

@@ -398,12 +398,11 @@ public interface SemanticNode {
      */
     default Map<Page, Rectangle2D> getBBox() {
 
-        Map<Page, Rectangle2D> bBoxPerPage = new HashMap<>();
         if (isLeaf()) {
-            return getBBoxFromLeafTextBlock(bBoxPerPage);
+            return getBBoxFromLeafTextBlock();
         }
 
-        return getBBoxFromChildren(bBoxPerPage);
+        return getBBoxFromChildren();
     }
 
 
@@ -426,25 +425,31 @@ public interface SemanticNode {
 
     /**
      * TODO: this produces unwanted results for sections spanning multiple columns.
-     *
-     * @param bBoxPerPage initial empty BoundingBox
+     * Computes the Union of the bounding boxes of all children recursively.
      * @return The union of the BoundingBoxes of all children
      */
-    private Map<Page, Rectangle2D> getBBoxFromChildren(Map<Page, Rectangle2D> bBoxPerPage) {
+    private Map<Page, Rectangle2D> getBBoxFromChildren() {
 
-        return streamChildren().map(SemanticNode::getBBox).reduce((map1, map2) -> {
-            map1.forEach((page, rectangle) -> map2.merge(page, rectangle, (rect1, rect2) -> rect1.createUnion(rect2).getBounds2D()));
-            return map2;
-        }).orElse(bBoxPerPage);
+        Map<Page, Rectangle2D> bBoxPerPage = new HashMap<>();
+        List<Map<Page, Rectangle2D>> childrenBBoxes = streamChildren().map(SemanticNode::getBBox).toList();
+        Set<Page> pages = childrenBBoxes.stream().flatMap(map -> map.keySet().stream()).collect(Collectors.toSet());
+        for (Page page : pages) {
+            Rectangle2D bBoxOnPage = childrenBBoxes.stream()
+                    .filter(childBboxPerPage -> childBboxPerPage.containsKey(page))
+                    .map(childBboxPerPage -> childBboxPerPage.get(page))
+                    .collect(RectangleTransformations.collectBBox());
+            bBoxPerPage.put(page, bBoxOnPage);
+        }
+        return bBoxPerPage;
     }
 
 
     /**
-     * @param bBoxPerPage initial empty BoundingBox
      * @return The union of all BoundingBoxes of the TextBlock of this node
      */
-    private Map<Page, Rectangle2D> getBBoxFromLeafTextBlock(Map<Page, Rectangle2D> bBoxPerPage) {
+    private Map<Page, Rectangle2D> getBBoxFromLeafTextBlock() {
 
+        Map<Page, Rectangle2D> bBoxPerPage = new HashMap<>();
         Map<Page, List<AtomicTextBlock>> atomicTextBlockPerPage = getTextBlock().getAtomicTextBlocks().stream().collect(Collectors.groupingBy(AtomicTextBlock::getPage));
         atomicTextBlockPerPage.forEach((page, atbs) -> bBoxPerPage.put(page, RectangleTransformations.bBoxUnionAtomicTextBlock(atbs)));
         return bBoxPerPage;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/graph/nodes/Table.java

@@ -2,10 +2,12 @@ package com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes;
 
 import static java.lang.String.format;
 
+import java.awt.geom.Rectangle2D;
 import java.util.Collection;
 import java.util.HashSet;
 import java.util.List;
 import java.util.Locale;
+import java.util.Map;
 import java.util.Set;
 import java.util.stream.IntStream;
 import java.util.stream.Stream;
@@ -40,6 +42,8 @@ public class Table implements SemanticNode {
     @EqualsAndHashCode.Exclude
     Set<RedactionEntity> entities = new HashSet<>();
 
+    @EqualsAndHashCode.Exclude
+    Map<Page, Rectangle2D> bBoxCache;
 
     /**
      * Streams all entities in this table, that appear in a row, which contains any of the provided strings.
@@ -311,5 +315,12 @@ public class Table implements SemanticNode {
 
         return treeId.toString() + ": " + NodeType.TABLE + ": #cols: " + numberOfCols + ", #rows: " + numberOfRows + ", " + this.getTextBlock().buildSummary();
     }
+    @Override
+    public Map<Page, Rectangle2D> getBBox() {
 
+        if (bBoxCache == null) {
+            bBoxCache = SemanticNode.super.getBBox();
+        }
+        return bBoxCache;
+    }
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/text/TextPositionSequence.java

@@ -9,8 +9,6 @@ import java.util.stream.Collectors;
 
 import org.apache.pdfbox.text.TextPosition;
 
-import com.dslplatform.json.JsonAttribute;
-import com.fasterxml.jackson.annotation.JsonIgnore;
 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Point;
 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
 
@@ -142,8 +140,7 @@ public class TextPositionSequence implements CharSequence {
      *
      * @return the text direction adjusted minX value
      */
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
+
     public float getMinXDirAdj() {
 
         return textPositions.get(0).getXDirAdj();
@@ -157,8 +154,7 @@ public class TextPositionSequence implements CharSequence {
      *
      * @return the text direction adjusted maxX value
      */
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
+
     public float getMaxXDirAdj() {
 
         return textPositions.get(textPositions.size() - 1).getXDirAdj() + textPositions.get(textPositions.size() - 1).getWidthDirAdj() + HEIGHT_PADDING;
@@ -172,8 +168,7 @@ public class TextPositionSequence implements CharSequence {
      *
      * @return the text direction adjusted minY value. The upper border of the bounding box of the word.
      */
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
+
     public float getMinYDirAdj() {
 
         return textPositions.get(0).getYDirAdj() - getTextHeight();
@@ -187,8 +182,7 @@ public class TextPositionSequence implements CharSequence {
      *
      * @return the text direction adjusted maxY value. The lower border of the bounding box of the word.
      */
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
+
     public float getMaxYDirAdj() {
 
         return textPositions.get(0).getYDirAdj();
@@ -196,38 +190,24 @@ public class TextPositionSequence implements CharSequence {
     }
 
 
-    public float getTextHeightNoPadding() {
-
-        return textPositions.get(0).getHeightDir();
-    }
-
-
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
     public float getTextHeight() {
 
         return textPositions.get(0).getHeightDir() + HEIGHT_PADDING;
     }
 
 
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
     public float getHeight() {
 
         return getMaxYDirAdj() - getMinYDirAdj();
     }
 
 
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
     public float getWidth() {
 
         return getMaxXDirAdj() - getMinXDirAdj();
     }
 
 
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
     public String getFont() {
 
         if (textPositions.get(0).getFontName() == null) {
@@ -237,8 +217,6 @@ public class TextPositionSequence implements CharSequence {
     }
 
 
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
     public String getFontStyle() {
 
         if (textPositions.get(0).getFontName() == null) {
@@ -258,16 +236,12 @@ public class TextPositionSequence implements CharSequence {
     }
 
 
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
     public float getFontSize() {
 
         return textPositions.get(0).getFontSizeInPt();
     }
 
 
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
     public float getSpaceWidth() {
 
         return textPositions.get(0).getWidthOfSpace();
@@ -283,8 +257,7 @@ public class TextPositionSequence implements CharSequence {
      *
      * @return bounding box of the word in Pdf Coordinate System
      */
-    @JsonIgnore
-    @JsonAttribute(ignore = true)
+
     @SneakyThrows
     public Rectangle getRectangle() {
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/blockification/DocuMineBlockificationService.java

@@ -61,8 +61,8 @@ public class DocuMineBlockificationService {
             boolean newLineAfterSplit = prev != null && word.getMinYDirAdj() != prev.getMinYDirAdj() && wasSplitted && splitX1 != word.getMinXDirAdj();
             boolean isSplitByRuling = isSplitByRuling(minX, minY, maxX, maxY, word, horizontalRulingLines, verticalRulingLines);
             boolean splitByDir = prev != null && !prev.getDir().equals(word.getDir());
-            boolean splitByOtherFontAndOtherY = prev != null && Math.abs(prev.getMaxYDirAdj() - word.getMaxYDirAdj()) > word.getTextHeight() * 0.2 && (word.getFontStyle()
-                    .contains("bold") && !prev.getFontStyle().contains("bold") || prev.getFontStyle().contains("bold") && !word.getFontStyle().contains("bold"));
+            boolean splitByOtherFontAndOtherY = prev != null && prev.getMaxYDirAdj() != word.getMaxYDirAdj() && (word.getFontStyle().contains("bold") && !prev.getFontStyle()
+                    .contains("bold") || prev.getFontStyle().contains("bold") && !word.getFontStyle().contains("bold"));
 
             Matcher matcher = pattern.matcher(chunkWords.stream().collect(Collectors.joining(" ")).toString());
             boolean startsOnSameX = Math.abs(minX - word.getMinXDirAdj()) < 5 && matcher.matches();

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/classification/DocuMineClassificationService.java

@@ -5,6 +5,7 @@ import java.util.Locale;
 import java.util.regex.Matcher;
 import java.util.regex.Pattern;
 
+import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 import org.springframework.stereotype.Service;
 
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
@@ -12,8 +13,6 @@ import com.knecon.fforesight.service.layoutparser.processor.model.Classification
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
 import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.utils.HeaderFooterDetection;
-import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 import com.knecon.fforesight.service.layoutparser.processor.utils.PositionUtils;
 
 import lombok.RequiredArgsConstructor;
@@ -24,7 +23,7 @@ import lombok.extern.slf4j.Slf4j;
 @RequiredArgsConstructor
 public class DocuMineClassificationService {
 
-    private static final Pattern pattern = Pattern.compile("^([1-9]\\d?\\.){1,3}\\d{1,2}\\.?\\s[0-9A-Za-z \\[\\]]{2,50}", Pattern.CASE_INSENSITIVE);
+    private static final Pattern pattern = Pattern.compile("^(\\d{1,2}\\.){1,3}\\d{1,2}\\.?\\s[0-9A-Za-z \\[\\]]{2,50}", Pattern.CASE_INSENSITIVE);
     private static final Pattern pattern2 = Pattern.compile("\\p{L}{3,}", Pattern.CASE_INSENSITIVE);
     private static final Pattern pattern3 = Pattern.compile("^(\\d{1,1}\\.){1,3}\\d{1,2}\\.?\\s[a-z]{1,2}\\/[a-z]{1,2}.*");
 
@@ -65,64 +64,46 @@ public class DocuMineClassificationService {
             return;
         }
         if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.HEADER)
-            || (PositionUtils.isOverBodyTextFrame(bodyTextFrame,
-                                                  textBlock,
-                                                  page.getRotation()) && (document.getFontSizeCounter().getMostPopular()
-                                                                          == null
-                                                                          || textBlock.getHighestFontSize()
-                                                                             <= document.getFontSizeCounter()
-                                                                                     .getMostPopular()))) {
+                || PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
+                .getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular())
+        ) {
             textBlock.setClassification(PageBlockType.HEADER);
 
         } else if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.FOOTER)
-                   || (PositionUtils.isUnderBodyTextFrame(bodyTextFrame,
-                                                          textBlock,
-                                                          page.getRotation())
-                       && (document.getFontSizeCounter().getMostPopular()
-                           == null
-                           || textBlock.getHighestFontSize()
-                              <= document.getFontSizeCounter()
-                                      .getMostPopular()))
-                   || HeaderFooterDetection.isLikelyFooter(textBlock, document, page)) {
+                || PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
+                .getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular())
+        ) {
             textBlock.setClassification(PageBlockType.FOOTER);
-        } else if (page.getPageNumber() == 1 && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock, document.getTextHeightCounter().getMostPopular()) > 2.5
-                                                 && textBlock.getHighestFontSize() > document.getFontSizeCounter().getMostPopular() || page.getTextBlocks().size() == 1)) {
+        } else if (page.getPageNumber() == 1 && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock,
+                document.getTextHeightCounter().getMostPopular()) > 2.5 && textBlock.getHighestFontSize() > document.getFontSizeCounter().getMostPopular() || page.getTextBlocks()
+                .size() == 1)) {
             if (!Pattern.matches("[0-9]+", textBlock.toString())) {
                 textBlock.setClassification(PageBlockType.TITLE);
             }
-        } else if (textBlock.getText().length() > 5
-                   && (textBlock.getMostPopularWordHeight() > document.getTextHeightCounter().getMostPopular()
-                       || textBlock.getMostPopularWordFontSize() > document.getFontSizeCounter().getMostPopular())
-                   && PositionUtils.getApproxLineCount(textBlock) < 5.9
-                   && (textBlock.getMostPopularWordStyle().contains("bold") && Character.isDigit(textBlock.toString().charAt(0)) && matcher2.reset().find() && !textBlock.toString()
-                .contains(":")
-                       || textBlock.toString().equals(textBlock.toString().toUpperCase(Locale.ROOT)) && matcher2.reset().find() && !textBlock.toString().contains(":")
-                       || textBlock.toString().startsWith("APPENDIX")
-                       || textBlock.toString().startsWith("FIGURE")
-                       || textBlock.toString().startsWith("TABLE"))
-                   && !textBlock.toString().endsWith(":")
-                   && matcher2.reset().find()) {
+        } else if (textBlock.getText().length() > 5 && (textBlock.getMostPopularWordHeight() > document.getTextHeightCounter()
+                .getMostPopular() || textBlock.getMostPopularWordFontSize() > document.getFontSizeCounter().getMostPopular()) && PositionUtils.getApproxLineCount(textBlock) < 5.9
+
+                && (textBlock.getMostPopularWordStyle().contains("bold") && Character.isDigit(textBlock.toString().charAt(0)) && !matcher2.matches() && !textBlock.toString()
+                .contains(":") || textBlock.toString().equals(textBlock.toString().toUpperCase(Locale.ROOT)) && !matcher2.matches() && !textBlock.toString()
+                .contains(":") || textBlock.toString().startsWith("APPENDIX") || textBlock.toString().startsWith("FIGURE") || textBlock.toString()
+                .startsWith("TABLE")) && !textBlock.toString().endsWith(":") && matcher2.find()) {
             textBlock.setClassification(PageBlockType.getHeadlineType(1));
             document.setHeadlines(true);
 
-        } else if (matcher.reset().find() && PositionUtils.getApproxLineCount(textBlock) < 2.9 && matcher2.reset().find() && !matcher3.reset().matches()) {
+        } else if (matcher.find() && PositionUtils.getApproxLineCount(textBlock) < 2.9 && matcher2.find() && !matcher3.matches()) {
             textBlock.setClassification(PageBlockType.getHeadlineType(2));
             document.setHeadlines(true);
-        } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock)
-                   && textBlock.getMostPopularWordFontSize() == document.getFontSizeCounter().getMostPopular()
-                   && textBlock.getMostPopularWordStyle().equals("bold")
-                   && !document.getFontStyleCounter().getMostPopular().equals("bold")) {
+        } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock) && textBlock.getMostPopularWordFontSize() == document.getFontSizeCounter()
+                .getMostPopular() && textBlock.getMostPopularWordStyle().equals("bold") && !document.getFontStyleCounter().getMostPopular().equals("bold")) {
             textBlock.setClassification(PageBlockType.PARAGRAPH_BOLD);
-        } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock)
-                   && textBlock.getMostPopularWordFont().equals(document.getFontCounter().getMostPopular())
-                   && textBlock.getMostPopularWordStyle().equals(document.getFontStyleCounter().getMostPopular())
-                   && textBlock.getMostPopularWordFontSize() == document.getFontSizeCounter().getMostPopular()) {
+        } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock) && textBlock.getMostPopularWordFont()
+                .equals(document.getFontCounter().getMostPopular()) && textBlock.getMostPopularWordStyle()
+                .equals(document.getFontStyleCounter().getMostPopular()) && textBlock.getMostPopularWordFontSize() == document.getFontSizeCounter().getMostPopular()) {
             textBlock.setClassification(PageBlockType.PARAGRAPH);
-        } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock)
-                   && textBlock.getMostPopularWordFontSize() == document.getFontSizeCounter().getMostPopular()
-                   && textBlock.getMostPopularWordStyle().equals("italic")
-                   && !document.getFontStyleCounter().getMostPopular().equals("italic")
-                   && PositionUtils.getApproxLineCount(textBlock) < 2.9) {
+        } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock) && textBlock.getMostPopularWordFontSize() == document.getFontSizeCounter()
+                .getMostPopular() && textBlock.getMostPopularWordStyle().equals("italic") && !document.getFontStyleCounter()
+                .getMostPopular()
+                .equals("italic") && PositionUtils.getApproxLineCount(textBlock) < 2.9) {
             textBlock.setClassification(PageBlockType.PARAGRAPH_ITALIC);
         } else if (PositionUtils.isWithinBodyTextFrame(bodyTextFrame, textBlock)) {
             textBlock.setClassification(PageBlockType.PARAGRAPH_UNKNOWN);

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/visualization/ViewerDocumentService.java

@@ -2,11 +2,14 @@ package com.knecon.fforesight.service.layoutparser.processor.services.visualizat
 
 import java.awt.geom.AffineTransform;
 import java.awt.geom.Rectangle2D;
+import java.io.File;
 import java.io.IOException;
-import java.io.OutputStream;
+import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.StandardCopyOption;
 
+import org.apache.pdfbox.Loader;
 import org.apache.pdfbox.cos.COSName;
-import org.apache.pdfbox.pdfwriter.compress.CompressParameters;
 import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDDocumentCatalog;
 import org.apache.pdfbox.pdmodel.PDPage;
@@ -29,6 +32,9 @@ import com.knecon.fforesight.service.layoutparser.processor.model.visualization.
 import com.knecon.fforesight.service.layoutparser.processor.model.visualization.PlacedText;
 import com.knecon.fforesight.service.layoutparser.processor.model.visualization.VisualizationsOnPage;
 
+import io.micrometer.observation.Observation;
+import io.micrometer.observation.ObservationRegistry;
+import io.micrometer.observation.annotation.Observed;
 import lombok.RequiredArgsConstructor;
 import lombok.SneakyThrows;
 import lombok.extern.slf4j.Slf4j;
@@ -43,29 +49,32 @@ public class ViewerDocumentService {
     public static final float LINE_WIDTH = 1f;
 
     private final LayoutGridService layoutGridService;
+    private final ObservationRegistry observationRegistry;
 
 
     @SneakyThrows
-    public void createViewerDocument(PDDocument pdDocument, Document document, OutputStream outputStream, boolean layerVisibilityDefaultValue) {
+    @Observed(name = "ViewerDocumentService", contextualName = "create-viewer-document")
+    public void createViewerDocument(File originFile, Document document, File destinationFile, boolean layerVisibilityDefaultValue) {
 
+        Path tmpFile = Files.createTempFile("tmpViewerDocument", ".pdf");
+        PDDocument pdDocument = openPDDocument(originFile);
         LayoutGrid layoutGrid = layoutGridService.createLayoutGrid(document);
-        // PDDocument.save() is very slow, since it actually traverses the entire pdf and writes a new one.
-        // If we collect all COSDictionaries we changed and tell it explicitly to only add the changed ones by using saveIncremental it's very fast.
+
         PDOptionalContentGroup layer = addLayerToDocument(pdDocument, layerVisibilityDefaultValue);
         PDFont font = new PDType1Font(Standard14Fonts.FontName.HELVETICA);
 
         for (int pageNumber = 0; pageNumber < pdDocument.getNumberOfPages(); pageNumber++) {
-            PDPage pdPage = pdDocument.getPage(pageNumber);
 
+            PDPage pdPage = pdDocument.getPage(pageNumber);
+//
             AffineTransform textDeRotationMatrix = getTextDeRotationTransform(pdPage);
-            addLayerToPageRessources(pdPage);
+            addLayerToPageResources(pdPage);
 
             // We need to save the graphics state before, such that our appended content cannot be affected by previous content streams with side effects,
             // e.g. not escaped matrix transformations.
             escapePreviousContents(pdDocument, pdPage);
 
-            VisualizationsOnPage visualizationsOnPage = layoutGrid.getVisualizationsPerPages()
-                    .get(pageNumber);
+            VisualizationsOnPage visualizationsOnPage = layoutGrid.getVisualizationsPerPages().get(pageNumber);
             assert pageNumber == visualizationsOnPage.getPageNumber();
             // We need to append to the content stream, otherwise the content could be overlapped by following content.
             try (var contentStream = new PDPageContentStream(pdDocument, pdPage, PDPageContentStream.AppendMode.APPEND, true)) {
@@ -99,11 +108,11 @@ public class ViewerDocumentService {
                     contentStream.setFont(font, FONT_SIZE);
                     contentStream.beginText();
                     Matrix textMatrix = new Matrix((float) textDeRotationMatrix.getScaleX(),
-                                                   (float) textDeRotationMatrix.getShearX(),
-                                                   (float) textDeRotationMatrix.getShearY(),
-                                                   (float) textDeRotationMatrix.getScaleY(),
-                                                   (float) placedText.lineStart().getX(),
-                                                   (float) placedText.lineStart().getY());
+                            (float) textDeRotationMatrix.getShearX(),
+                            (float) textDeRotationMatrix.getShearY(),
+                            (float) textDeRotationMatrix.getScaleY(),
+                            (float) placedText.lineStart().getX(),
+                            (float) placedText.lineStart().getY());
                     textMatrix.translate(-((font.getStringWidth(placedText.text()) / 1000) * FONT_SIZE + (2 * LINE_WIDTH) + 4), -FONT_SIZE);
                     contentStream.setTextMatrix(textMatrix);
                     contentStream.showText(placedText.text());
@@ -112,13 +121,48 @@ public class ViewerDocumentService {
                 contentStream.restoreGraphicsState();
                 contentStream.endMarkedContent();
             }
+
+            if (pageNumber % 500 == 0 && pageNumber != 0) { // re-open document every once in a while to save on RAM
+                log.info("Incremental save after {} pages", pageNumber);
+                observedIncrementalSave(pdDocument, destinationFile);
+                pdDocument.close();
+                Files.copy(destinationFile.toPath(), tmpFile, StandardCopyOption.REPLACE_EXISTING);
+                pdDocument = openPDDocument(tmpFile.toFile());
+                layer = addLayerToDocument(pdDocument, layerVisibilityDefaultValue);
+            }
+
         }
-//        dictionariesToUpdate.add(pdDocument.getDocument().getTrailer());
-        pdDocument.save(outputStream, CompressParameters.NO_COMPRESSION);
+        observedIncrementalSave(pdDocument, destinationFile);
+
+        tmpFile.toFile().delete();
+        pdDocument.close();
     }
 
 
-    private static void addLayerToPageRessources(PDPage pdPage) {
+    private static PDDocument openPDDocument(File tmpFile) throws IOException {
+
+        PDDocument pdDocument;
+        pdDocument = Loader.loadPDF(tmpFile);
+        pdDocument.setAllSecurityToBeRemoved(true);
+        return pdDocument;
+    }
+
+
+    @SneakyThrows
+    private void observedIncrementalSave(PDDocument pdDocument, File outputFile) {
+
+        Observation.createNotStarted("ViewerDocumentService", observationRegistry).contextualName("incremental-save").observe(() -> {
+            try {
+                pdDocument.save(outputFile);
+            } catch (IOException e) {
+                throw new RuntimeException(e);
+            }
+        });
+
+    }
+
+
+    private static void addLayerToPageResources(PDPage pdPage) {
 
         PDResources resources = pdPage.getResources();
         if (resources == null) {
@@ -155,6 +199,7 @@ public class ViewerDocumentService {
             ocprops.addGroup(layer);
         }
         ocprops.setGroupEnabled(layer, layerVisibilityDefaultValue);
+//        dictionariesToUpdate.add(catalog.getCOSObject());
         return layer;
     }
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/utils/HeaderFooterDetection.java

@@ -1,223 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.utils;
-
-import java.util.ArrayList;
-import java.util.HashMap;
-import java.util.List;
-import java.util.Map;
-import java.util.stream.Collectors;
-
-import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
-import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
-
-import lombok.experimental.UtilityClass;
-
-@UtilityClass
-public class HeaderFooterDetection {
-
-    private final Map<Integer, ClassificationPage> pagesCache = new HashMap<>();
-    private static final double THRESHOLD = 0.5;
-    // Weight will go from 1.0 to 0.5 because the first element is the most likely to be the header on the page.
-    private static final double[] headerWeights = {1.0, 0.75, 0.5};
-    // Weight will go from 0.5 to 1.0 because the last element is the most likely to be the footer on the page.
-    private static final double[] footerWeights = {0.5, 0.75, 1.0};
-
-
-    public boolean isLikelyFooter(TextPageBlock textPageBlock, ClassificationDocument document, ClassificationPage classificationPage) {
-
-        int numberOfPages = document.getPages().size();
-        if (numberOfPages < 3) {
-            // If the document has 1 or 2 pages this may lead to more false positives than actual findings.
-            return false;
-        }
-
-        int window = Math.min(numberOfPages, 8);
-
-        List<ClassificationPage> nearestPages = findNearestPages(classificationPage, document.getPages(), window);
-        List<List<AbstractPageBlock>> footerCandidates = getFooterCandidates(nearestPages);
-
-        return detectHeadersOrFootersByPageAssociation(textPageBlock.getText(), footerCandidates, window, footerWeights);
-    }
-
-
-    public boolean isLikelyHeader(TextPageBlock textPageBlock, ClassificationDocument document, ClassificationPage classificationPage) {
-
-        int numberOfPages = document.getPages().size();
-        if (numberOfPages < 3) {
-            // If the document has 1 or 2 pages this may lead to more false positives than actual findings.
-            return false;
-        }
-
-        int window = Math.min(numberOfPages, 8);
-
-        List<ClassificationPage> nearestPages = findNearestPages(classificationPage, document.getPages(), window);
-        List<List<AbstractPageBlock>> headerCandidates = getHeaderCandidates(nearestPages);
-
-        return detectHeadersOrFootersByPageAssociation(textPageBlock.getText(), headerCandidates, window, headerWeights);
-    }
-
-
-    private boolean detectHeadersOrFootersByPageAssociation(String testString, List<List<AbstractPageBlock>> candidates, int window, double[] weights) {
-
-        double highestScore = 0.0;
-
-        for (int i = 0; i < candidates.size(); i++) {
-            List<List<String>> candidateStrings = new ArrayList<>();
-            for (int k = Math.max(i - window, 0); k < Math.min(i + window, candidates.size()); k++) {
-                candidateStrings.add(candidates.get(k)
-                                             .stream()
-                                             .map(AbstractPageBlock::getText)
-                                             .collect(Collectors.toList()));
-            }
-
-            int maxLen = candidateStrings.stream()
-                    .mapToInt(List::size)
-                    .max()
-                    .orElse(0);
-            for (List<String> sublist : candidateStrings) {
-                while (sublist.size() < maxLen) {
-                    sublist.add(0, "");
-                }
-            }
-
-            // Compare the testString against each candidate in the window
-            for (int j = 0; j < maxLen; j++) {
-                double score = 0.0;
-                int finalJ = j;
-                List<String> paddedCandidateStrings = candidateStrings.stream()
-                        .map(sublist -> sublist.size() > finalJ ? sublist.get(finalJ) : "")
-                        .toList();
-                for (String paddedString : paddedCandidateStrings) {
-                    if ((testString.length() >= 5 && paddedString.length() >= 5) && (testString.length() > 2 * paddedString.length()
-                                                                                     || paddedString.length() > 2 * testString.length())) {
-                        // If both strings are at least 5 characters long and one string is more than twice as long as the other,
-                        // skip the distance calculation as it's time-consuming, and we can assume they are not similar enough
-                        continue;
-                    }
-
-                    int distance = calculateHammingDistanceWithPreprocessing(testString, paddedString);
-                    double normalizedScore = 1 - (double) distance / Math.max(testString.length(), paddedString.length());
-                    score += normalizedScore * (j < weights.length ? weights[j] : 1);
-                }
-                score /= paddedCandidateStrings.size();
-                highestScore = Math.max(highestScore, score);
-                // Early stop
-                if (highestScore > THRESHOLD) {
-                    return true;
-                }
-            }
-        }
-
-        return false;
-    }
-
-
-    /**
-     * Find the nearest n pages for a given page.
-     * For example: nearest 8 pages for page 4 are: 1, 2, 3, 5, 6, 7, 8, 9.
-     *
-     * @param currentPage  Current page to find the nearest ones.
-     * @param allPages     All pages in the document.
-     * @param numNeighbors Number of neighbouring pages to find.
-     * @return The nearest pages.
-     */
-    private List<ClassificationPage> findNearestPages(ClassificationPage currentPage, List<ClassificationPage> allPages, int numNeighbors) {
-
-        int totalPages = allPages.size();
-        List<ClassificationPage> nearestPages = new ArrayList<>();
-
-        int currentPageIndex = currentPage.getPageNumber() - 1;
-        int halfWin = numNeighbors / 2;
-        int start = Math.max(0, currentPageIndex - halfWin);
-        int end = Math.min(totalPages - 1, currentPageIndex + halfWin);
-
-        for (int i = start; i <= end; i++) {
-            if (i != currentPageIndex) {
-                nearestPages.add(pagesCache.computeIfAbsent(i, allPages::get));
-            }
-        }
-
-        pagesCache.keySet().removeIf(key -> key < start || key > end);
-
-        return nearestPages;
-    }
-
-
-    // Get the last 3 TextBlocks on the page as they are likely to be a footer
-    private List<List<AbstractPageBlock>> getFooterCandidates(List<ClassificationPage> pages) {
-
-        List<List<AbstractPageBlock>> footerCandidates = new ArrayList<>();
-        for (ClassificationPage page : pages) {
-            List<AbstractPageBlock> textBlocks = page.getTextBlocks();
-            List<TextPageBlock> textPageBlocks = textBlocks.stream()
-                    .filter(textBlock -> textBlock instanceof TextPageBlock)
-                    .map(textBlock -> (TextPageBlock) textBlock)
-                    .toList();
-            int blockCount = textPageBlocks.size();
-            if (blockCount > 0) {
-                int start = Math.max(0, blockCount - 3);
-                footerCandidates.add(new ArrayList<>(textPageBlocks.subList(start, blockCount)));
-            }
-        }
-        return footerCandidates;
-    }
-
-
-    // Get the first 3 TextBlocks on the page as they are likely to be a header
-    private List<List<AbstractPageBlock>> getHeaderCandidates(List<ClassificationPage> pages) {
-
-        List<List<AbstractPageBlock>> headerCandidates = new ArrayList<>();
-        for (ClassificationPage page : pages) {
-            List<AbstractPageBlock> textBlocks = page.getTextBlocks();
-            List<TextPageBlock> textPageBlocks = textBlocks.stream()
-                    .filter(textBlock -> textBlock instanceof TextPageBlock)
-                    .map(textBlock -> (TextPageBlock) textBlock)
-                    .toList();
-            int count = Math.min(3, textPageBlocks.size());
-            headerCandidates.add(new ArrayList<>(textPageBlocks.subList(0, count)));
-        }
-        return headerCandidates;
-    }
-
-
-    /**
-     * Calculate the Hamming distance between two strings after preprocessing to make them the same length
-     * and replacing all digits with a special character '@' since they are a common occurrence in headers/footers.
-     *
-     * @param firstCandidate  First string
-     * @param secondCandidate Second string
-     * @return The Hamming distance between the two preprocessed strings.
-     */
-    private int calculateHammingDistanceWithPreprocessing(String firstCandidate, String secondCandidate) {
-
-        int maxLength = Math.max(firstCandidate.length(), secondCandidate.length());
-
-        String cleanFirstCandidate = padString(firstCandidate, maxLength, '\0').replaceAll("\\d", "@");
-        String cleanSecondCandidate = padString(secondCandidate, maxLength, '\0').replaceAll("\\d", "@");
-
-        int distance = 0;
-        for (int i = 0; i < maxLength; i++) {
-            if (cleanFirstCandidate.charAt(i) != cleanSecondCandidate.charAt(i)) {
-                distance++;
-            }
-        }
-        return distance;
-    }
-
-
-    private String padString(String input, int length, char padChar) {
-
-        if (input.length() >= length) {
-            return input;
-        }
-
-        StringBuilder sb = new StringBuilder(input);
-
-        while (sb.length() < length) {
-            sb.append(padChar);
-        }
-        return sb.toString();
-    }
-
-}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/utils/TextPositionSequenceComparator.java

@@ -28,13 +28,15 @@ import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPosit
  *
  * @author Ben Litchfield
  */
-public class TextPositionSequenceComparator implements Comparator<TextPositionSequence> {
-
+public class TextPositionSequenceComparator implements Comparator<TextPositionSequence>
+{
     @Override
-    public int compare(TextPositionSequence pos1, TextPositionSequence pos2) {
+    public int compare(TextPositionSequence pos1, TextPositionSequence pos2)
+    {
         // only compare text that is in the same direction
         int cmp1 = Float.compare(pos1.getDir().getDegrees(), pos2.getDir().getDegrees());
-        if (cmp1 != 0) {
+        if (cmp1 != 0)
+        {
             return cmp1;
         }
 
@@ -46,19 +48,25 @@ public class TextPositionSequenceComparator implements Comparator<TextPositionSe
         float pos2YBottom = pos2.getMaxYDirAdj();
 
         // note that the coordinates have been adjusted so 0,0 is in upper left
-        float pos1YTop = pos1YBottom - pos1.getTextHeightNoPadding();
-        float pos2YTop = pos2YBottom - pos2.getTextHeightNoPadding();
+        float pos1YTop = pos1YBottom - pos1.getTextHeight();
+        float pos2YTop = pos2YBottom - pos2.getTextHeight();
 
         float yDifference = Math.abs(pos1YBottom - pos2YBottom);
 
         // we will do a simple tolerance comparison
-        if (yDifference < .1 || pos2YBottom >= pos1YTop && pos2YBottom <= pos1YBottom || pos1YBottom >= pos2YTop && pos1YBottom <= pos2YBottom) {
+        if (yDifference < .1 ||
+                pos2YBottom >= pos1YTop && pos2YBottom <= pos1YBottom ||
+                pos1YBottom >= pos2YTop && pos1YBottom <= pos2YBottom)
+        {
             return Float.compare(x1, x2);
-        } else if (pos1YBottom < pos2YBottom) {
+        }
+        else if (pos1YBottom < pos2YBottom)
+        {
             return -1;
-        } else {
+        }
+        else
+        {
             return 1;
         }
     }
-
 }

layoutparser-service/layoutparser-service-server/build.gradle.kts

@@ -3,40 +3,52 @@ import org.springframework.boot.gradle.tasks.bundling.BootBuildImage
 plugins {
     application
     id("com.knecon.fforesight.java-conventions")
-    id("org.springframework.boot") version "3.1.3"
+    id("org.springframework.boot") version "3.1.5"
     id("io.spring.dependency-management") version "1.1.3"
     id("org.sonarqube") version "4.3.0.3225"
-    id("io.freefair.lombok") version "8.6"
+    id("io.freefair.lombok") version "8.2.2"
 //    id("org.graalvm.buildtools.native") version "0.9.23"
 }
 
 description = "layoutparser-service-server"
 
+configurations {
+    all {
+        exclude(group = "commons-logging", module = "commons-logging")
+        exclude(group = "org.springframework.boot", module = "spring-boot-starter-log4j2")
+        exclude(group = "com.iqser.red.commons", module = "logging-commons")
+    }
+}
+
+
 val jacksonVersion = "2.15.2"
 val pdfBoxVersion = "3.0.0"
+val springBootStarterVersion = "3.1.5"
 
 dependencies {
     implementation(project(":layoutparser-service-processor"))
     implementation(project(":layoutparser-service-internal-api"))
 
     implementation("com.iqser.red.commons:storage-commons:2.45.0")
-    implementation("com.knecon.fforesight:tenant-commons:0.14.0")
-    implementation("org.springframework.boot:spring-boot-starter-actuator:3.1.3")
+    implementation("com.knecon.fforesight:tenant-commons:0.19.0")
+    implementation("com.knecon.fforesight:tracing-commons:0.3.0")
+    implementation("org.springframework.boot:spring-boot-starter-actuator:${springBootStarterVersion}")
+    implementation("org.springframework.boot:spring-boot-starter-amqp:${springBootStarterVersion}")
     implementation("com.amazonaws:aws-java-sdk-s3:1.12.536")
-    implementation("org.springframework.boot:spring-boot-starter-amqp:3.1.3")
     implementation("org.springframework.cloud:spring-cloud-starter-openfeign:4.0.4")
+    implementation("net.logstash.logback:logstash-logback-encoder:7.4")
 
     // for integration testing only
     testImplementation(project(":layoutparser-service-internal-api"))
 
-    testImplementation("org.springframework.boot:spring-boot-starter-amqp:3.1.3")
+    testImplementation("org.springframework.boot:spring-boot-starter-amqp:${springBootStarterVersion}")
+    testImplementation("org.springframework.boot:spring-boot-starter-test:${springBootStarterVersion}")
     testImplementation("com.iqser.red.service:persistence-service-shared-api-v1:2.144.0")
     testImplementation("com.iqser.red.commons:jackson-commons:1.3.0")
     testImplementation("com.fasterxml.jackson.module:jackson-module-afterburner:${jacksonVersion}")
     testImplementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:${jacksonVersion}")
     testImplementation("org.apache.pdfbox:pdfbox:${pdfBoxVersion}")
     testImplementation("org.apache.pdfbox:pdfbox-tools:${pdfBoxVersion}")
-    testImplementation("org.springframework.boot:spring-boot-starter-test:3.1.3")
 }
 
 /*

layoutparser-service/layoutparser-service-server/src/main/java/com/knecon/fforesight/service/layoutparser/server/Application.java

@@ -5,6 +5,7 @@ import org.springframework.boot.actuate.autoconfigure.security.servlet.Managemen
 import org.springframework.boot.autoconfigure.ImportAutoConfiguration;
 import org.springframework.boot.autoconfigure.SpringBootApplication;
 import org.springframework.boot.autoconfigure.security.servlet.SecurityAutoConfiguration;
+import org.springframework.context.annotation.Bean;
 import org.springframework.context.annotation.Import;
 
 import com.amazonaws.services.s3.model.metrics.MetricsConfiguration;
@@ -13,8 +14,11 @@ import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingService
 import com.knecon.fforesight.service.layoutparser.server.queue.MessagingConfiguration;
 import com.knecon.fforesight.tenantcommons.MultiTenancyAutoConfiguration;
 
+import io.micrometer.observation.ObservationRegistry;
+import io.micrometer.observation.aop.ObservedAspect;
+
 @ImportAutoConfiguration({MultiTenancyAutoConfiguration.class})
-@Import({MetricsConfiguration.class, StorageAutoConfiguration.class,  LayoutParsingServiceProcessorConfiguration.class, MessagingConfiguration.class})
+@Import({MetricsConfiguration.class, StorageAutoConfiguration.class, LayoutParsingServiceProcessorConfiguration.class, MessagingConfiguration.class})
 @SpringBootApplication(exclude = {SecurityAutoConfiguration.class, ManagementWebSecurityAutoConfiguration.class})
 public class Application {
 
@@ -23,4 +27,11 @@ public class Application {
         SpringApplication.run(Application.class, args);
     }
 
+
+    @Bean
+    public ObservedAspect observedAspect(ObservationRegistry observationRegistry) {
+
+        return new ObservedAspect(observationRegistry);
+    }
+
 }

layoutparser-service/layoutparser-service-server/src/main/java/com/knecon/fforesight/service/layoutparser/server/queue/MessagingConfiguration.java

@@ -24,8 +24,7 @@ public class MessagingConfiguration {
     @Bean
     public Queue layoutparsingResponseQueue() {
 
-        return QueueBuilder.durable(LayoutParsingQueueNames.LAYOUT_PARSING_FINISHED_EVENT_QUEUE)//
-                .withArgument("x-dead-letter-exchange", "").withArgument("x-dead-letter-routing-key", LayoutParsingQueueNames.LAYOUT_PARSING_DLQ).build();
+        return QueueBuilder.durable(LayoutParsingQueueNames.LAYOUT_PARSING_FINISHED_EVENT_QUEUE).build();
     }
 
 

layoutparser-service/layoutparser-service-server/src/main/resources/application.yml

@@ -5,10 +5,19 @@ tenant-user-management-service.url: "http://tenant-user-management-service:8080/
 
 fforesight.tenants.remote: true
 
+logging.pattern.level: "%5p [${spring.application.name},%X{traceId:-},%X{spanId:-}]"
+
+logging.type: ${LOGGING_TYPE:CONSOLE}
+kubernetes.namespace: ${NAMESPACE:default}
+project.version: 1.0-SNAPSHOT
+
+
 server:
   port: 8080
 
 spring:
+  application:
+    name: layoutparser-service
   main:
     allow-circular-references: true # FIXME
   rabbitmq:
@@ -32,6 +41,13 @@ management:
     prometheus.enabled: ${monitoring.enabled:false}
     health.enabled: true
   endpoints.web.exposure.include: prometheus, health
+  tracing:
+    enabled: ${TRACING_ENABLED:false}
+    sampling:
+      probability: ${TRACING_PROBABILITY:1.0}
+  otlp:
+    tracing:
+      endpoint: ${OTLP_ENDPOINT:http://otel-collector-opentelemetry-collector.otel-collector:4318/v1/traces}
 
 
 storage:

layoutparser-service/layoutparser-service-server/src/main/resources/logback-spring.xml

@@ -0,0 +1,17 @@
+<configuration>
+
+    <springProperty scope="configuration" name="logType" source="logging.type"/>
+    <springProperty scope="context" name="application.name" source="spring.application.name"/>
+    <springProperty scope="context" name="version" source="project.version"/>
+    <include resource="org/springframework/boot/logging/logback/defaults.xml"/>
+    <include resource="org/springframework/boot/logging/logback/console-appender.xml"/>
+
+    <appender name="JSON" class="ch.qos.logback.core.ConsoleAppender">
+        <encoder class="net.logstash.logback.encoder.LogstashEncoder"/>
+    </appender>
+
+    <root level="INFO">
+        <appender-ref ref="${logType}"/>
+    </root>
+
+</configuration>

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/BdrJsonBuildTest.java

@@ -46,12 +46,11 @@ public class BdrJsonBuildTest extends AbstractTest {
     @SneakyThrows
     protected Document buildGraph(File file) {
 
-        try (PDDocument pdDocument = Loader.loadPDF(file)) {
-            return DocumentGraphFactory.buildDocumentGraph(layoutParsingPipeline.parseLayout(LayoutParsingType.TAAS,
-                    pdDocument,
-                    new ImageServiceResponse(),
-                    new TableServiceResponse()));
-        }
+        return DocumentGraphFactory.buildDocumentGraph(layoutParsingPipeline.parseLayout(LayoutParsingType.TAAS,
+                file,
+                new ImageServiceResponse(),
+                new TableServiceResponse(),
+                file.toString()));
     }
 
 

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/HeadlinesGoldStandardIntegrationTest.java

@@ -6,7 +6,6 @@ import java.util.List;
 import java.util.Set;
 import java.util.stream.Collectors;
 
-import org.apache.pdfbox.Loader;
 import org.assertj.core.api.Assertions;
 import org.junit.jupiter.api.Test;
 import org.junit.jupiter.api.extension.ExtendWith;
@@ -96,9 +95,10 @@ public class HeadlinesGoldStandardIntegrationTest {
         goldStandardLog.getRedactionLogEntry().forEach(e -> goldStandardHeadlines.add(new Headline(e.getPositions().get(0).getPage(), e.getValue())));
 
         Document documentGraph = DocumentGraphFactory.buildDocumentGraph(layoutParsingPipeline.parseLayout(LayoutParsingType.REDACT_MANAGER,
-                Loader.loadPDF(pdfFileResource.getFile()),
+                pdfFileResource.getFile(),
                 new ImageServiceResponse(),
-                new TableServiceResponse()));
+                new TableServiceResponse(),
+                filePath));
 
         var foundHeadlines = documentGraph.streamAllSubNodes()
                 .map(SemanticNode::getHeadline)

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/LayoutparserEnd2EndTest.java

@@ -25,8 +25,8 @@ public class LayoutparserEnd2EndTest extends AbstractTest {
     @SneakyThrows
     public void testLayoutParserEndToEnd() {
 
-        prepareStorage("files/bdr/btd_19_053_1905391.pdf");
-        LayoutParsingRequest layoutParsingRequest = buildDefaultLayoutParsingRequest(LayoutParsingType.TAAS);
+        prepareStorage("files/bdr/Wie weiter bei Kristeneinrichtungen.pdf");
+        LayoutParsingRequest layoutParsingRequest = buildDefaultLayoutParsingRequest(LayoutParsingType.REDACT_MANAGER);
         LayoutParsingFinishedEvent finishedEvent = layoutParsingPipeline.parseLayoutAndSaveFilesToStorage(layoutParsingRequest);
         Arrays.stream(finishedEvent.message().split("\n")).forEach(log::info);
     }

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/graph/DocumentDataTests.java

@@ -6,6 +6,7 @@ import java.nio.file.Files;
 import java.nio.file.Path;
 import java.util.List;
 
+import org.junit.jupiter.api.Disabled;
 import org.junit.jupiter.api.Test;
 import org.springframework.core.io.ClassPathResource;
 
@@ -17,7 +18,9 @@ import com.knecon.fforesight.service.layoutparser.server.utils.BuildDocumentTest
 import lombok.SneakyThrows;
 
 public class DocumentDataTests extends BuildDocumentTest {
+
     @Test
+    @Disabled // This test takes waaaaaay too long, it's ridiculous
     @SneakyThrows
     public void createDocumentDataForAllFiles() {
 
@@ -36,11 +39,12 @@ public class DocumentDataTests extends BuildDocumentTest {
         for (String pdfFileName : pdfFileNames) {
             System.out.println(pdfFileName);
             DocumentData documentData = DocumentDataMapper.toDocumentData(buildGraph(resource.getFile().toPath().getParent().relativize(Path.of(pdfFileName)).toString()));
-                        File outputFile = Path.of(outPath).resolve(resource.getFile().toPath().relativize(Path.of(pdfFileName))).toFile();
+            File outputFile = Path.of(outPath).resolve(resource.getFile().toPath().relativize(Path.of(pdfFileName))).toFile();
             outputFile.toPath().getParent().toFile().mkdirs();
             try (var out = new FileOutputStream(outputFile.toString().replace(".pdf", ".json"))) {
                 ObjectMapperFactory.create().writeValue(out, documentData);
             }
         }
     }
+
 }

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/graph/DocumentGraphJsonWritingTest.java

@@ -5,7 +5,6 @@ import java.io.FileOutputStream;
 import java.nio.file.Files;
 import java.nio.file.Path;
 
-import org.apache.pdfbox.Loader;
 import org.junit.jupiter.api.Disabled;
 import org.junit.jupiter.api.Test;
 import org.springframework.core.io.ClassPathResource;
@@ -56,9 +55,10 @@ public class DocumentGraphJsonWritingTest extends BuildDocumentTest {
     private void writeJsons(Path filename) {
 
         Document documentGraph = DocumentGraphFactory.buildDocumentGraph(layoutParsingPipeline.parseLayout(LayoutParsingType.REDACT_MANAGER,
-                Loader.loadPDF(filename.toFile()),
+                filename.toFile(),
                 new ImageServiceResponse(),
-                new TableServiceResponse()));
+                new TableServiceResponse(),
+                filename.toFile().toString()));
 
         DocumentData documentData = DocumentDataMapper.toDocumentData(documentGraph);
         ObjectMapper mapper = ObjectMapperFactory.create();

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/graph/ViewerDocumentTest.java

@@ -1,32 +1,13 @@
 package com.knecon.fforesight.service.layoutparser.server.graph;
 
-import java.io.FileOutputStream;
+import java.io.File;
 import java.nio.file.Path;
-import java.util.List;
 
-import org.apache.pdfbox.Loader;
-import org.apache.pdfbox.pdmodel.PDDocument;
-import org.junit.jupiter.api.Disabled;
 import org.junit.jupiter.api.Test;
-import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.core.io.ClassPathResource;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructure;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
-import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Table;
-import com.knecon.fforesight.service.layoutparser.processor.model.table.Cell;
-import com.knecon.fforesight.service.layoutparser.processor.model.table.TablePageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.python_api.model.image.ImageServiceResponse;
-import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
-import com.knecon.fforesight.service.layoutparser.processor.services.SectionsBuilderService;
-import com.knecon.fforesight.service.layoutparser.processor.services.classification.RedactManagerClassificationService;
-import com.knecon.fforesight.service.layoutparser.processor.services.factory.DocumentGraphFactory;
-import com.knecon.fforesight.service.layoutparser.processor.services.mapper.DocumentDataMapper;
-import com.knecon.fforesight.service.layoutparser.processor.services.mapper.PropertiesMapper;
 import com.knecon.fforesight.service.layoutparser.processor.services.visualization.LayoutGridService;
 import com.knecon.fforesight.service.layoutparser.processor.services.visualization.ViewerDocumentService;
 import com.knecon.fforesight.service.layoutparser.server.utils.BuildDocumentTest;
@@ -35,38 +16,17 @@ import lombok.SneakyThrows;
 
 public class ViewerDocumentTest extends BuildDocumentTest {
 
-    @Autowired
-    private SectionsBuilderService sectionsBuilderService;
-
-    @Autowired
-    private RedactManagerClassificationService redactManagerClassificationService;
-
     @Test
     @SneakyThrows
     public void testViewerDocument() {
 
-        String fileName = "files/bdr/notMergedParagraphs.pdf";
+        String fileName = "files/S-Metolachlor_RAR_01_Volume_1_2018-09-06.pdf";
         String tmpFileName = "/tmp/" + Path.of(fileName).getFileName() + "_VIEWER.pdf";
         LayoutGridService layoutGridService = new LayoutGridService();
-        ViewerDocumentService viewerDocumentService = new ViewerDocumentService(layoutGridService);
-        Document document = buildGraph(fileName, LayoutParsingType.TAAS);
-        try (var pdDocument = Loader.loadPDF(new ClassPathResource(fileName).getFile()); var out = new FileOutputStream(tmpFileName)) {
-            viewerDocumentService.createViewerDocument(pdDocument, document, out, true);
-        }
-    }
-
-    public ClassificationDocument buildClassificationDocument(PDDocument originDocument) {
-
-        ClassificationDocument classificationDocument = layoutParsingPipeline.parseLayout(LayoutParsingType.REDACT_MANAGER,
-                originDocument,
-                new ImageServiceResponse(),
-                new TableServiceResponse());
-
-        redactManagerClassificationService.classifyDocument(classificationDocument);
-
-        sectionsBuilderService.buildSections(classificationDocument);
-
-        return classificationDocument;
+        ViewerDocumentService viewerDocumentService = new ViewerDocumentService(layoutGridService, null);
+        Document document = buildGraph(fileName, LayoutParsingType.REDACT_MANAGER);
+        var documentFile = new ClassPathResource(fileName).getFile();
+        viewerDocumentService.createViewerDocument(documentFile, document, new File(tmpFileName), true);
     }
 
 }

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/segmentation/PdfSegmentationServiceTest.java

@@ -3,6 +3,7 @@ package com.knecon.fforesight.service.layoutparser.server.segmentation;
 import static org.assertj.core.api.Assertions.assertThat;
 
 import java.awt.geom.Rectangle2D;
+import java.io.File;
 import java.io.FileOutputStream;
 import java.io.IOException;
 import java.nio.file.Path;
@@ -15,8 +16,6 @@ import java.util.Locale;
 import java.util.Map;
 import java.util.stream.Collectors;
 
-import org.apache.pdfbox.Loader;
-import org.apache.pdfbox.pdmodel.PDDocument;
 import org.junit.jupiter.api.Disabled;
 import org.junit.jupiter.api.Test;
 import org.springframework.beans.factory.annotation.Autowired;
@@ -62,12 +61,13 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
 
     @SneakyThrows
-    public ClassificationDocument buildClassificationDocument(PDDocument originDocument, TableServiceResponse tableServiceResponse) {
+    public ClassificationDocument buildClassificationDocument(File originDocument, TableServiceResponse tableServiceResponse) {
 
         ClassificationDocument classificationDocument = layoutParsingPipeline.parseLayout(LayoutParsingType.REDACT_MANAGER,
                 originDocument,
                 new ImageServiceResponse(),
-                tableServiceResponse);
+                tableServiceResponse,
+                "document");
 
         redactManagerClassificationService.classifyDocument(classificationDocument);
 
@@ -78,7 +78,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
 
     @SneakyThrows
-    public ClassificationDocument buildClassificationDocument(PDDocument originDocument) {
+    public ClassificationDocument buildClassificationDocument(File originDocument) {
 
         return buildClassificationDocument(originDocument, new TableServiceResponse());
     }
@@ -89,7 +89,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/A20622A izRMS (CZ) fRR Part B9_Page185.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         toHtml(document, "/tmp/A20622A izRMS (CZ) fRR Part B9_Page185.html");
 
@@ -103,7 +103,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
         ClassPathResource cvTablesResource = new ClassPathResource("files/cv_tables/ScanRotationBorder.TABLES.json");
         var tableServiceResponse = objectMapper.readValue(cvTablesResource.getInputStream(), TableServiceResponse.class);
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()), tableServiceResponse);
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile(), tableServiceResponse);
 
         toHtml(document, "/tmp/ScanRotationBorder.html");
     }
@@ -116,7 +116,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
         ClassPathResource cvTablesResource = new ClassPathResource("files/cv_tables/ScanRotationBorder.TABLES.json");
         var tableServiceResponse = objectMapper.readValue(cvTablesResource.getInputStream(), TableServiceResponse.class);
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()), tableServiceResponse);
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile(), tableServiceResponse);
         assertThat(document.getSections().stream().flatMap(paragraph -> paragraph.getTables().stream()).collect(Collectors.toList())).isNotEmpty();
         var tables = document.getSections().stream().flatMap(paragraph -> paragraph.getTables().stream()).toList();
 
@@ -156,7 +156,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/Minimal Examples/Spanning Cells.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
         assertThat(document.getSections().stream().flatMap(paragraph -> paragraph.getTables().stream()).collect(Collectors.toList())).isNotEmpty();
         TablePageBlock table = document.getSections().stream().flatMap(paragraph -> paragraph.getTables().stream()).collect(Collectors.toList()).get(0);
         assertThat(table.getColCount()).isEqualTo(6);
@@ -170,7 +170,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/Minimal Examples/Merge Table.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
         assertThat(document.getSections().stream().flatMap(paragraph -> paragraph.getTables().stream()).collect(Collectors.toList())).isNotEmpty();
         TablePageBlock firstTable = document.getSections().stream().flatMap(paragraph -> paragraph.getTables().stream()).collect(Collectors.toList()).get(0);
         assertThat(firstTable.getColCount()).isEqualTo(8);
@@ -188,7 +188,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/Minimal Examples/Merge Multi Page Table.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
         assertThat(document.getSections().stream().flatMap(paragraph -> paragraph.getTables().stream()).collect(Collectors.toList())).isNotEmpty();
         TablePageBlock firstTable = document.getSections().stream().flatMap(paragraph -> paragraph.getTables().stream()).collect(Collectors.toList()).get(0);
         assertThat(firstTable.getColCount()).isEqualTo(9);
@@ -206,7 +206,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/Minimal Examples/Rotated Table Headers.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
         assertThat(document.getSections().stream().flatMap(paragraph -> paragraph.getTables().stream()).collect(Collectors.toList())).isNotEmpty();
         TablePageBlock firstTable = document.getSections().stream().flatMap(paragraph -> paragraph.getTables().stream()).collect(Collectors.toList()).get(0);
         assertThat(firstTable.getColCount()).isEqualTo(8);
@@ -224,7 +224,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/56 Fludioxonil_RAR_12_Volume_3CA_B-7_2018-02-21_Page170.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 4);
 
@@ -241,7 +241,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/211.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 4);
 
@@ -258,7 +258,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/VV-931175_Page1.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 1);
 
@@ -299,7 +299,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/27 A8637C - EU AIR3 - MCP Section 1 - Identity of the plant protection product_Page6.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         toHtml(document, "/tmp/html.html");
 
@@ -319,7 +319,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/A20622A izRMS (CZ) fRR Part B9_Page185.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 1);
 
@@ -332,7 +332,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/A20622A izRMS (CZ) fRR Part B9_Page185_fixed.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 1);
 
@@ -345,7 +345,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/A20622A izZRMS (CZ) fRR Part B7_Page123.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 6);
 
@@ -364,7 +364,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/77 Pirimicarb_RAR_08_Volume_3CA_B-6_2017-12-04_Page11.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 3);
 
@@ -380,7 +380,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/95 Trinexapac-ethyl_RAR_08_Volume_3CA_B-6_2018-01-10_Page532.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 1);
         validateTable(document, 0, 9, 9, 0, 0);
@@ -393,7 +393,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/52 Fludioxonil_RAR_07_Volume_3CA_B-5_2018-02-21_Page175.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 1);
 
@@ -407,7 +407,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/52 Fludioxonil_RAR_07_Volume_3CA_B-5_2018-02-21_Page174.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 1);
         validateTable(document, 0, 9, 6, 7, 0);
@@ -420,7 +420,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/19 Chlorothalonil RAR 08 Volume 3CA B 6b metabolites Oct 2017_Page35.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 1);
         validateTable(document, 0, 10, 6, 0, 0);
@@ -433,7 +433,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/19 Chlorothalonil RAR 08 Volume 3CA B 6b metabolites Oct 2017_Page161.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 2);
         validateTable(document, 0, 2, 2, 0, 0);
@@ -448,7 +448,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
         ClassPathResource pdfFileResource = new ClassPathResource(
                 "files/SinglePages/47 Cyprodinil - EU AIR3 - MCA Section 5 Supplement - Toxicological and metabolism studies on the active substance_Page30.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 2);
 
@@ -464,7 +464,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
         ClassPathResource pdfFileResource = new ClassPathResource(
                 "files/SinglePages/49 Cyprodinil - EU AIR3 - MCA Section 8 Supplement - Ecotoxicological studies on the active substance_Page61.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 2);
 
@@ -479,7 +479,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/81 Pirimicarb_RAR_20_Volume_3CP_A10788A (_Pirimor_)_B-9_2017-12-04_Page54.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 2);
 
@@ -494,7 +494,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/85 Pydiflumetofen_DAR_08_Volume_3CA_B-6_2017-07-26_Page134.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 2);
 
@@ -509,7 +509,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/Thiabendazole DAR Addendum for ED_April_2020_Page18.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 4);
 
@@ -526,7 +526,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/15 - Pretilachlor - Acute Oral Toxicity (Up and Down Procedure) - Rat_Page18.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 1);
 
@@ -541,7 +541,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
         ClassPathResource pdfFileResource = new ClassPathResource(
                 "files/SinglePages/28 A8637C - EU AIR3 - MCP Section 10 - Ecotoxicological studies on the plant protection product_Page23.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 2);
 
@@ -556,7 +556,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/24 - SYN549522 - Acute Oral Toxicity - Rats_Page17.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 1);
 
@@ -570,7 +570,7 @@ public class PdfSegmentationServiceTest extends AbstractTest {
 
         ClassPathResource pdfFileResource = new ClassPathResource("files/SinglePages/30 - Dicamba - Acute Oral Toxicity - Rats_Page5.pdf");
 
-        ClassificationDocument document = buildClassificationDocument(Loader.loadPDF(pdfFileResource.getFile()));
+        ClassificationDocument document = buildClassificationDocument(pdfFileResource.getFile());
 
         validateTableSize(document, 1);
 

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/services/RulingCleaningServiceTest.java

@@ -58,7 +58,7 @@ public class RulingCleaningServiceTest extends BuildDocumentTest {
     public void testTableExtraction() {
 
         LayoutGridService layoutGridService = new LayoutGridService();
-        ViewerDocumentService viewerDocumentService = new ViewerDocumentService(layoutGridService);
+        ViewerDocumentService viewerDocumentService = new ViewerDocumentService(layoutGridService, null);
 
         ClassPathResource resource = new ClassPathResource("files");
         List<String> pdfFileNames = Files.walk(resource.getFile().toPath())
@@ -77,13 +77,15 @@ public class RulingCleaningServiceTest extends BuildDocumentTest {
     private void writeJsons(Path filename) {
 
         Document documentGraphBefore = DocumentGraphFactory.buildDocumentGraph(layoutParsingPipeline.parseLayout(LayoutParsingType.REDACT_MANAGER,
-                Loader.loadPDF(filename.toFile()),
+                filename.toFile(),
                 new ImageServiceResponse(),
-                new TableServiceResponse()));
+                new TableServiceResponse(),
+                filename.toFile().toString()));
         Document documentGraphAfter = DocumentGraphFactory.buildDocumentGraph(layoutParsingPipeline.parseLayout(LayoutParsingType.REDACT_MANAGER,
-                Loader.loadPDF(filename.toFile()),
+                filename.toFile(),
                 new ImageServiceResponse(),
-                new TableServiceResponse()));
+                new TableServiceResponse(),
+                filename.toFile().toString()));
         DocumentData documentDataBefore = DocumentDataMapper.toDocumentData(documentGraphBefore);
         DocumentData documentDataAfter = DocumentDataMapper.toDocumentData(documentGraphAfter);
         if (!compareStructures(documentDataBefore.getDocumentStructure(), documentDataAfter.getDocumentStructure(), filename.getFileName().toString())) {

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/utils/AbstractTest.java

@@ -1,5 +1,26 @@
 package com.knecon.fforesight.service.layoutparser.server.utils;
 
+import java.io.InputStream;
+import java.util.Map;
+import java.util.Optional;
+
+import org.junit.jupiter.api.AfterEach;
+import org.junit.jupiter.api.BeforeEach;
+import org.junit.jupiter.api.extension.ExtendWith;
+import org.springframework.amqp.rabbit.core.RabbitTemplate;
+import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.boot.autoconfigure.EnableAutoConfiguration;
+import org.springframework.boot.autoconfigure.amqp.RabbitAutoConfiguration;
+import org.springframework.boot.test.context.SpringBootTest;
+import org.springframework.boot.test.mock.mockito.MockBean;
+import org.springframework.context.annotation.Bean;
+import org.springframework.context.annotation.ComponentScan;
+import org.springframework.context.annotation.Configuration;
+import org.springframework.context.annotation.Import;
+import org.springframework.context.annotation.Primary;
+import org.springframework.core.io.ClassPathResource;
+import org.springframework.test.context.junit.jupiter.SpringExtension;
+
 import com.iqser.red.commons.jackson.ObjectMapperFactory;
 import com.iqser.red.storage.commons.service.StorageService;
 import com.iqser.red.storage.commons.utils.FileSystemBackedStorageService;
@@ -9,22 +30,8 @@ import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingStorage
 import com.knecon.fforesight.service.layoutparser.server.Application;
 import com.knecon.fforesight.tenantcommons.TenantContext;
 import com.knecon.fforesight.tenantcommons.TenantsClient;
-import lombok.SneakyThrows;
-import org.junit.jupiter.api.AfterEach;
-import org.junit.jupiter.api.BeforeEach;
-import org.junit.jupiter.api.extension.ExtendWith;
-import org.springframework.amqp.rabbit.core.RabbitTemplate;
-import org.springframework.beans.factory.annotation.Autowired;
-import org.springframework.boot.autoconfigure.EnableAutoConfiguration;
-import org.springframework.boot.autoconfigure.amqp.RabbitAutoConfiguration;
-import org.springframework.boot.test.context.SpringBootTest;
-import org.springframework.boot.test.mock.mockito.MockBean;
-import org.springframework.context.annotation.*;
-import org.springframework.core.io.ClassPathResource;
-import org.springframework.test.context.junit.jupiter.SpringExtension;
 
-import java.io.InputStream;
-import java.util.Optional;
+import lombok.SneakyThrows;
 
 @ExtendWith(SpringExtension.class)
 @SpringBootTest(classes = Application.class, webEnvironment = SpringBootTest.WebEnvironment.RANDOM_PORT)
@@ -100,9 +107,11 @@ public abstract class AbstractTest {
         return buildDefaultLayoutParsingRequest(LayoutParsingType.REDACT_MANAGER);
     }
 
+
     protected LayoutParsingRequest buildDefaultLayoutParsingRequest(LayoutParsingType layoutParsingType) {
 
         return LayoutParsingRequest.builder()
+                .identifier(Map.of("fileId", "1337"))
                 .layoutParsingType(layoutParsingType)
                 .originFileStorageId(ORIGIN_FILE_ID)
                 .tablesFileStorageId(Optional.of(TABLE_FILE_ID))
@@ -116,6 +125,7 @@ public abstract class AbstractTest {
                 .build();
     }
 
+
     @SneakyThrows
     protected LayoutParsingRequest prepareStorage(String file, String cvServiceResponseFile, String imageInfoFile) {
 
@@ -152,7 +162,6 @@ public abstract class AbstractTest {
     @ComponentScan("com.knecon.fforesight.service.layoutparser")
     public static class TestConfiguration {
 
-
         @Bean
         @Primary
         public StorageService inmemoryStorage() {

layoutparser-service/layoutparser-service-server/src/test/java/com/knecon/fforesight/service/layoutparser/server/utils/BuildDocumentTest.java

@@ -1,9 +1,7 @@
 package com.knecon.fforesight.service.layoutparser.server.utils;
 
-import java.io.InputStream;
+import java.io.File;
 
-import org.apache.pdfbox.Loader;
-import org.apache.pdfbox.pdmodel.PDDocument;
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.core.io.ClassPathResource;
 
@@ -25,11 +23,9 @@ public abstract class BuildDocumentTest extends AbstractTest {
     @SneakyThrows
     protected ClassificationDocument parseLayout(String filename, LayoutParsingType layoutParsingType) {
 
-        ClassPathResource fileResource = new ClassPathResource(filename);
+        File fileResource = new ClassPathResource(filename).getFile();
         prepareStorage(filename);
-        try (InputStream inputStream = fileResource.getInputStream(); PDDocument pdDocument = Loader.loadPDF(inputStream.readAllBytes())) {
-            return layoutParsingPipeline.parseLayout(layoutParsingType, pdDocument, layoutParsingStorageService.getImagesFile(IMAGE_FILE_ID), new TableServiceResponse());
-        }
+        return layoutParsingPipeline.parseLayout(layoutParsingType, fileResource, layoutParsingStorageService.getImagesFile(IMAGE_FILE_ID), new TableServiceResponse(), filename);
     }
 
 

layoutparser-service/layoutparser-service-server/src/test/resources/application.yml

@@ -7,6 +7,9 @@ fforesight.tenants.remote: true
 server:
   port: 8080
 
+logging.pattern.level: "%5p [${spring.application.name},%X{traceId:-},%X{spanId:-}]"
+logging.type: ${LOGGING_TYPE:CONSOLE}
+
 spring:
   main:
     allow-circular-references: true # FIXME

layoutparser-service/layoutparser-service-server/src/test/resources/log4j2-test.xml

@@ -1,16 +0,0 @@
-<Configuration>
-
-    <Appenders>
-        <Console name="CONSOLE" target="SYSTEM_OUT">
-            <PatternLayout pattern="%d{HH:mm:ss.SSS} [%t] %-5level %logger{36} - %msg%n"/>
-        </Console>
-    </Appenders>
-
-    <Loggers>
-        <Root level="warn">
-            <AppenderRef ref="CONSOLE"/>
-        </Root>
-        <Logger name="com.iqser" level="info"/>
-    </Loggers>
-
-</Configuration>

layoutparser-service/layoutparser-service-server/src/test/resources/logback-spring.xml

@@ -0,0 +1,17 @@
+<configuration>
+
+    <springProperty scope="configuration" name="logType" source="logging.type"/>
+    <springProperty scope="context" name="application.name" source="spring.application.name"/>
+    <springProperty scope="context" name="version" source="project.version"/>
+    <include resource="org/springframework/boot/logging/logback/defaults.xml"/>
+    <include resource="org/springframework/boot/logging/logback/console-appender.xml"/>
+
+    <appender name="JSON" class="ch.qos.logback.core.ConsoleAppender">
+        <encoder class="net.logstash.logback.encoder.LogstashEncoder"/>
+    </appender>
+
+    <root level="INFO">
+        <appender-ref ref="${logType}"/>
+    </root>
+
+</configuration>

publish-custom-image.sh

@@ -2,7 +2,14 @@
 dir=${PWD##*/}
 gradle assemble
 
-buildNumber=${1:-1}
+# Get the current Git branch
+branch=$(git rev-parse --abbrev-ref HEAD)
 
-gradle bootBuildImage --cleanCache --publishImage -PbuildbootDockerHostNetwork=true -Pversion=$USER-$buildNumber
-echo "nexus.knecon.com:5001/red/${dir}-server-v1:$USER-$buildNumber"
+# Get the short commit hash (first 5 characters)
+commit_hash=$(git rev-parse --short=5 HEAD)
+
+# Combine branch and commit hash
+buildName="${USER}-${branch}-${commit_hash}"
+
+gradle bootBuildImage --publishImage -PbuildbootDockerHostNetwork=true -Pversion=$buildName --no-build-cache
+echo "nexus.knecon.com:5001/ff/${dir}-service-server:$buildName"