Pull request #316 : RED-2440: Integrated image-service-v2

Merge in RED/redaction-service from RED-2440 to master * commit 'dee1aa1f01085e585bb3126b7883bd3d01090f17': RED-2440: Integrated image-service-v2
RED-2440: Integrated image-service-v2
2021-12-22 10:40:04 +01:00 · 2021-12-22 10:35:13 +01:00 · 2021-12-21 12:17:40 +01:00 · 2021-12-21 12:00:24 +01:00 · 2021-12-21 11:07:37 +01:00 · 2021-12-21 11:18:37 +02:00
61 changed files with 4253 additions and 552 deletions
@@ -6,6 +6,7 @@
        <groupId>com.iqser.red</groupId>
        <artifactId>platform-docker-dependency</artifactId>
        <version>1.1.0</version>
+        <relativePath />
    </parent>
    <modelVersion>4.0.0</modelVersion>

@@ -5,7 +5,8 @@
    <parent>
        <artifactId>platform-dependency</artifactId>
        <groupId>com.iqser.red</groupId>
-        <version>1.1.5</version>
+        <version>1.5.0</version>
+        <relativePath />
    </parent>
    <modelVersion>4.0.0</modelVersion>

@@ -12,7 +12,7 @@
    <artifactId>redaction-service-api-v1</artifactId>

    <properties>
-        <persistence-service.version>0.121.0</persistence-service.version>
+        <persistence-service.version>0.151.0</persistence-service.version>
    </properties>

    <dependencies>
@@ -19,10 +19,11 @@ import java.util.Set;
@AllArgsConstructor
 public class AnalyzeRequest {

+    private MessageType messageType;
+
    private String dossierId;
    private String fileId;
    private String dossierTemplateId;
-    private boolean reanalyseOnlyIfPossible;
    private ManualRedactions manualRedactions;
    private OffsetDateTime lastProcessed;

@@ -1,5 +1,7 @@
 package com.iqser.red.service.redaction.v1.model;

+import com.iqser.red.service.persistence.service.v1.api.model.annotations.ManualRedactions;
+
 import lombok.AllArgsConstructor;
 import lombok.Builder;
 import lombok.Data;
@@ -11,6 +13,8 @@ import lombok.NoArgsConstructor;
@AllArgsConstructor
 public class AnalyzeResult {

+    private MessageType messageType;
+
    private String dossierId;
    private String fileId;
    private long duration;
@@ -25,6 +29,8 @@ public class AnalyzeResult {

    private int analysisVersion;

+    private ManualRedactions manualRedactions;
+

 }

@@ -0,0 +1,7 @@
+package com.iqser.red.service.redaction.v1.model;
+
+public enum MessageType {
+
+    FULL_ANALYSE, REANALYSE, SURROUNDING_TEXT
+
+}
@@ -30,4 +30,9 @@ public class SectionArea {
        return page == other.getPage() && this.topLeft.getX() <= other.getTopLeft().getX() && this.topLeft.getX() + this.getWidth() >= other.getTopLeft().getX() + other.getWidth() && this.getTopLeft().getY() <= other.getTopLeft().getY() && this.getTopLeft().getY() + this.getHeight() >= other.getTopLeft().getY() + other.getHeight();
    }

+    // TODO we should only use one rectangle class.
+    public boolean contains(com.iqser.red.service.persistence.service.v1.api.model.annotations.Rectangle other) {
+        return page == other.getPage() && this.topLeft.getX() <= other.getTopLeftX() && this.topLeft.getX() + this.getWidth() >= other.getTopLeftX() + other.getWidth() && this.getTopLeft().getY() <= other.getTopLeftY() && this.getTopLeft().getY() + this.getHeight() >= other.getTopLeftY() + other.getHeight();
+    }
+
 }
@@ -1,29 +1,43 @@
 package com.iqser.red.service.redaction.v1.resources;

+import com.iqser.red.service.persistence.service.v1.api.model.annotations.ManualRedactions;
 import com.iqser.red.service.redaction.v1.model.*;
+
 import org.springframework.http.MediaType;
 import org.springframework.web.bind.annotation.PathVariable;
 import org.springframework.web.bind.annotation.PostMapping;
 import org.springframework.web.bind.annotation.RequestBody;
+import org.springframework.web.bind.annotation.RequestParam;

 public interface RedactionResource {

    @PostMapping(value = "/annotate", produces = MediaType.APPLICATION_JSON_VALUE, consumes = MediaType.APPLICATION_JSON_VALUE)
    AnnotateResponse annotate(@RequestBody AnnotateRequest annotateRequest);

+
    @PostMapping(value = "/debug/classifications", produces = MediaType.APPLICATION_JSON_VALUE, consumes = MediaType.APPLICATION_JSON_VALUE)
    RedactionResult classify(@RequestBody RedactionRequest redactionRequest);

+
    @PostMapping(value = "/debug/sections", produces = MediaType.APPLICATION_JSON_VALUE, consumes = MediaType.APPLICATION_JSON_VALUE)
    RedactionResult sections(@RequestBody RedactionRequest redactionRequest);

+
    @PostMapping(value = "/debug/htmlTables", produces = MediaType.APPLICATION_JSON_VALUE, consumes = MediaType.APPLICATION_JSON_VALUE)
    RedactionResult htmlTables(@RequestBody RedactionRequest redactionRequest);

+
    @PostMapping(value = "/rules/test", consumes = MediaType.APPLICATION_JSON_VALUE)
    void testRules(@RequestBody String rules);

+
    @PostMapping(value = "/redaction-log/preview", consumes = MediaType.APPLICATION_JSON_VALUE)
    RedactionLog getRedactionLog(@RequestBody RedactionRequest redactionRequest);

+
+    @PostMapping(value = "/manual/surrounding-text/{dossierId}/{fileId}", consumes = MediaType.APPLICATION_JSON_VALUE, produces = MediaType.APPLICATION_JSON_VALUE)
+    ManualRedactions addSurroundingText(@PathVariable("dossierId") String dossierId,
+                                        @PathVariable("fileId") String fileId,
+                                        @RequestBody ManualRedactions manualRedactions);
+
 }
@@ -7,6 +7,7 @@ import lombok.Data;
 import lombok.NonNull;
 import lombok.RequiredArgsConstructor;

+import java.util.ArrayList;
 import java.util.List;

@Data
@@ -16,7 +17,7 @@ public class Page {
    @NonNull
    private List<AbstractTextContainer> textBlocks;

-    private List<PdfImage> images;
+    private List<PdfImage> images = new ArrayList<>();

    private Rectangle bodyTextFrame;

@@ -1,15 +0,0 @@
-package com.iqser.red.service.redaction.v1.server.client;
-
-import org.springframework.cloud.openfeign.FeignClient;
-import org.springframework.http.MediaType;
-import org.springframework.web.bind.annotation.PostMapping;
-import org.springframework.web.bind.annotation.RequestBody;
-import org.springframework.web.multipart.MultipartFile;
-
-@FeignClient(name = "ImageClassificationResource", url = "${image-service.url}")
-public interface ImageClassificationClient {
-
-    @PostMapping(value = "/process_full_img", consumes = MediaType.MULTIPART_FORM_DATA_VALUE, produces = MediaType.APPLICATION_JSON_VALUE)
-    ImageClassificationResponse classify(@RequestBody MultipartFile file);
-
-}
@@ -1,13 +0,0 @@
-package com.iqser.red.service.redaction.v1.server.client;
-
-import lombok.AllArgsConstructor;
-import lombok.Data;
-import lombok.NoArgsConstructor;
-
-@Data
-@NoArgsConstructor
-@AllArgsConstructor
-public class ImageClassificationResponse {
-
-    private String category;
-}
@@ -1,7 +1,24 @@
 package com.iqser.red.service.redaction.v1.server.controller;

+import java.io.ByteArrayOutputStream;
+import java.io.IOException;
+import java.util.stream.Collectors;
+
+import org.apache.pdfbox.io.MemoryUsageSetting;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.springframework.web.bind.annotation.PathVariable;
+import org.springframework.web.bind.annotation.RequestBody;
+import org.springframework.web.bind.annotation.RestController;
+
+import com.iqser.red.service.persistence.service.v1.api.model.annotations.ManualRedactions;
 import com.iqser.red.service.persistence.service.v1.api.model.dossiertemplate.dossier.file.FileType;
-import com.iqser.red.service.redaction.v1.model.*;
+import com.iqser.red.service.redaction.v1.model.AnnotateRequest;
+import com.iqser.red.service.redaction.v1.model.AnnotateResponse;
+import com.iqser.red.service.redaction.v1.model.RedactionLog;
+import com.iqser.red.service.redaction.v1.model.RedactionRequest;
+import com.iqser.red.service.redaction.v1.model.RedactionResult;
+import com.iqser.red.service.redaction.v1.model.SectionArea;
+import com.iqser.red.service.redaction.v1.model.SectionGrid;
 import com.iqser.red.service.redaction.v1.resources.RedactionResource;
 import com.iqser.red.service.redaction.v1.server.classification.model.Document;
 import com.iqser.red.service.redaction.v1.server.classification.model.Page;
@@ -10,22 +27,16 @@ import com.iqser.red.service.redaction.v1.server.exception.RedactionException;
 import com.iqser.red.service.redaction.v1.server.redaction.service.AnnotationService;
 import com.iqser.red.service.redaction.v1.server.redaction.service.DictionaryService;
 import com.iqser.red.service.redaction.v1.server.redaction.service.DroolsExecutionService;
+import com.iqser.red.service.redaction.v1.server.redaction.service.ManualRedactionSurroundingTextService;
 import com.iqser.red.service.redaction.v1.server.redaction.service.RedactionLogMergeService;
 import com.iqser.red.service.redaction.v1.server.segmentation.PdfSegmentationService;
 import com.iqser.red.service.redaction.v1.server.storage.RedactionStorageService;
 import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;
 import com.iqser.red.service.redaction.v1.server.tableextraction.model.Table;
 import com.iqser.red.service.redaction.v1.server.visualization.service.PdfVisualisationService;
+
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
-import org.apache.pdfbox.io.MemoryUsageSetting;
-import org.apache.pdfbox.pdmodel.PDDocument;
-import org.springframework.web.bind.annotation.RequestBody;
-import org.springframework.web.bind.annotation.RestController;
-
-import java.io.ByteArrayOutputStream;
-import java.io.IOException;
-import java.util.stream.Collectors;

@Slf4j
@RestController
@@ -39,11 +50,19 @@ public class RedactionController implements RedactionResource {
    private final PdfSegmentationService pdfSegmentationService;
    private final RedactionStorageService redactionStorageService;
    private final RedactionLogMergeService redactionLogMergeService;
+    private final ManualRedactionSurroundingTextService manualRedactionSurroundingTextService;
+

    public AnnotateResponse annotate(@RequestBody AnnotateRequest annotateRequest) {

-        var storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(annotateRequest.getDossierId(), annotateRequest.getFileId(), FileType.ORIGIN));
-        var mergedRedactionLog = getRedactionLog(RedactionRequest.builder().fileId(annotateRequest.getFileId()).manualRedactions(annotateRequest.getManualRedactions()).dossierId(annotateRequest.getDossierId()).dossierTemplateId(annotateRequest.getDossierTemplateId()).build());
+        var storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(annotateRequest
+                .getDossierId(), annotateRequest.getFileId(), FileType.ORIGIN));
+        var mergedRedactionLog = getRedactionLog(RedactionRequest.builder()
+                .fileId(annotateRequest.getFileId())
+                .manualRedactions(annotateRequest.getManualRedactions())
+                .dossierId(annotateRequest.getDossierId())
+                .dossierTemplateId(annotateRequest.getDossierTemplateId())
+                .build());
        var sectionsGrid = redactionStorageService.getSectionGrid(annotateRequest.getDossierId(), annotateRequest.getFileId());

        try (PDDocument pdDocument = PDDocument.load(storedObjectStream, MemoryUsageSetting.setupTempFileOnly())) {
@@ -65,11 +84,14 @@ public class RedactionController implements RedactionResource {

    @Override
    public RedactionResult classify(@RequestBody RedactionRequest redactionRequest) {
-        var storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(redactionRequest.getDossierId(), redactionRequest.getFileId(), FileType.ORIGIN));
-        try {
-            Document classifiedDoc = pdfSegmentationService.parseDocument(storedObjectStream);

-            storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(redactionRequest.getDossierId(), redactionRequest.getFileId(), FileType.ORIGIN));
+        var storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(redactionRequest
+                .getDossierId(), redactionRequest.getFileId(), FileType.ORIGIN));
+        try {
+            Document classifiedDoc = pdfSegmentationService.parseDocument(storedObjectStream, null);
+
+            storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(redactionRequest
+                    .getDossierId(), redactionRequest.getFileId(), FileType.ORIGIN));
            try (PDDocument pdDocument = PDDocument.load(storedObjectStream)) {
                pdDocument.setAllSecurityToBeRemoved(true);

@@ -85,16 +107,19 @@ public class RedactionController implements RedactionResource {
            throw new RedactionException(e);
        }

-
    }

+
    @Override
    public RedactionResult sections(@RequestBody RedactionRequest redactionRequest) {
-        var storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(redactionRequest.getDossierId(), redactionRequest.getFileId(), FileType.ORIGIN));
-        try {
-            Document classifiedDoc = pdfSegmentationService.parseDocument(storedObjectStream);

-            storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(redactionRequest.getDossierId(), redactionRequest.getFileId(), FileType.ORIGIN));
+        var storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(redactionRequest
+                .getDossierId(), redactionRequest.getFileId(), FileType.ORIGIN));
+        try {
+            Document classifiedDoc = pdfSegmentationService.parseDocument(storedObjectStream, null);
+
+            storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(redactionRequest
+                    .getDossierId(), redactionRequest.getFileId(), FileType.ORIGIN));
            try (PDDocument pdDocument = PDDocument.load(storedObjectStream)) {
                pdDocument.setAllSecurityToBeRemoved(true);

@@ -109,7 +134,6 @@ public class RedactionController implements RedactionResource {
            throw new RedactionException(e);
        }

-
    }


@@ -119,13 +143,13 @@ public class RedactionController implements RedactionResource {
        Document classifiedDoc;

        try {
-            var storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(redactionRequest.getDossierId(), redactionRequest.getFileId(), FileType.ORIGIN));
-            classifiedDoc = pdfSegmentationService.parseDocument(storedObjectStream, true);
+            var storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(redactionRequest
+                    .getDossierId(), redactionRequest.getFileId(), FileType.ORIGIN));
+            classifiedDoc = pdfSegmentationService.parseDocument(storedObjectStream, null);
        } catch (Exception e) {
            throw new RedactionException(e);
        }

-
        StringBuilder sb = new StringBuilder();
        for (Page page : classifiedDoc.getPages()) {
            for (AbstractTextContainer textContainer : page.getTextBlocks()) {
@@ -140,12 +164,14 @@ public class RedactionController implements RedactionResource {

    }

+
    @Override
    public void testRules(@RequestBody String rules) {

        droolsExecutionService.testRules(rules);
    }

+
    @Override
    public RedactionLog getRedactionLog(RedactionRequest redactionRequest) {

@@ -160,7 +186,8 @@ public class RedactionController implements RedactionResource {

        log.info("Loaded redaction log with computationalVersion: {}", redactionLog.getAnalysisVersion());

-        SectionGrid sectionGrid = redactionStorageService.getSectionGrid(redactionRequest.getDossierId(), redactionRequest.getFileId());
+        SectionGrid sectionGrid = redactionStorageService.getSectionGrid(redactionRequest.getDossierId(), redactionRequest
+                .getFileId());
        if (sectionGrid.getSections().isEmpty()) {

            log.info("SectionGrid does not have headlines set. Computing headlines now!");
@@ -168,24 +195,27 @@ public class RedactionController implements RedactionResource {

            // enhance section grid with headline data
            for (var sectionText : text.getSectionTexts()) {
-                sectionGrid.getSections().add(new SectionGrid.SectionGridSection(sectionText.getSectionNumber(),
-                        sectionText.getHeadline(),
-                        sectionText.getSectionAreas().stream().map(SectionArea::getPage).collect(Collectors.toSet()),
-                        sectionText.getSectionAreas()));
+                sectionGrid.getSections()
+                        .add(new SectionGrid.SectionGridSection(sectionText.getSectionNumber(), sectionText.getHeadline(), sectionText
+                                .getSectionAreas()
+                                .stream()
+                                .map(SectionArea::getPage)
+                                .collect(Collectors.toSet()), sectionText.getSectionAreas()));
            }
            redactionStorageService.storeObject(redactionRequest.getDossierId(), redactionRequest.getFileId(), FileType.SECTION_GRID, sectionGrid);
        }

-
        log.info("Loaded redaction log with computationalVersion: {}", redactionLog.getAnalysisVersion());
        if (redactionLog.getAnalysisVersion() == 0) {
            // old redaction logs are returned directly
            return redactionLog;
        } else {
-            return redactionLogMergeService.mergeRedactionLogData(redactionLog, sectionGrid, redactionRequest.getDossierTemplateId(), redactionRequest.getManualRedactions(), redactionRequest.getExcludedPages());
+            return redactionLogMergeService.mergeRedactionLogData(redactionLog, sectionGrid, redactionRequest.getDossierTemplateId(), redactionRequest
+                    .getManualRedactions(), redactionRequest.getExcludedPages());
        }
    }

+
    private RedactionResult convert(PDDocument document, int numberOfPages) throws IOException {

        try (ByteArrayOutputStream byteArrayOutputStream = new ByteArrayOutputStream()) {
@@ -199,4 +229,16 @@ public class RedactionController implements RedactionResource {
    }


+    @Override
+    public ManualRedactions addSurroundingText(@PathVariable("dossierId") String dossierId,
+                                               @PathVariable("fileId") String fileId,
+                                               @RequestBody ManualRedactions manualRedactions) {
+
+        long start = System.currentTimeMillis();
+        var result = manualRedactionSurroundingTextService.addSurroundingText(dossierId, fileId, manualRedactions);
+        log.info("add surrounding text for manual redaction in dossierId {} and fileId {} took: {}", dossierId, fileId, System
+                .currentTimeMillis() - start);
+        return result;
+    }
+
 }
@@ -0,0 +1,377 @@
+/*
+ * Licensed to the Apache Software Foundation (ASF) under one or more
+ * contributor license agreements.  See the NOTICE file distributed with
+ * this work for additional information regarding copyright ownership.
+ * The ASF licenses this file to You under the Apache License, Version 2.0
+ * (the "License"); you may not use this file except in compliance with
+ * the License.  You may obtain a copy of the License at
+ *
+ *      http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package com.iqser.red.service.redaction.v1.server.parsing;
+
+import java.io.InputStream;
+import java.io.IOException;
+import java.util.Map;
+import java.util.WeakHashMap;
+
+import org.apache.commons.logging.Log;
+import org.apache.commons.logging.LogFactory;
+
+import org.apache.fontbox.ttf.TrueTypeFont;
+import org.apache.fontbox.util.BoundingBox;
+
+import org.apache.pdfbox.contentstream.PDFStreamEngine;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.font.encoding.GlyphList;
+import org.apache.pdfbox.pdmodel.common.PDRectangle;
+import org.apache.pdfbox.pdmodel.font.PDCIDFont;
+import org.apache.pdfbox.pdmodel.font.PDCIDFontType2;
+import org.apache.pdfbox.pdmodel.font.PDFont;
+import org.apache.pdfbox.pdmodel.font.PDSimpleFont;
+import org.apache.pdfbox.pdmodel.font.PDTrueTypeFont;
+import org.apache.pdfbox.pdmodel.font.PDType0Font;
+import org.apache.pdfbox.pdmodel.font.PDType3Font;
+import org.apache.pdfbox.pdmodel.graphics.state.PDGraphicsState;
+import org.apache.pdfbox.text.TextPosition;
+import org.apache.pdfbox.util.Matrix;
+import org.apache.pdfbox.util.Vector;
+import org.apache.pdfbox.contentstream.operator.DrawObject;
+import org.apache.pdfbox.contentstream.operator.state.Concatenate;
+import org.apache.pdfbox.contentstream.operator.state.Restore;
+import org.apache.pdfbox.contentstream.operator.state.Save;
+import org.apache.pdfbox.contentstream.operator.state.SetGraphicsStateParameters;
+import org.apache.pdfbox.contentstream.operator.state.SetMatrix;
+import org.apache.pdfbox.contentstream.operator.text.BeginText;
+import org.apache.pdfbox.contentstream.operator.text.EndText;
+import org.apache.pdfbox.contentstream.operator.text.SetFontAndSize;
+import org.apache.pdfbox.contentstream.operator.text.SetTextHorizontalScaling;
+import org.apache.pdfbox.contentstream.operator.text.ShowTextAdjusted;
+import org.apache.pdfbox.contentstream.operator.text.ShowTextLine;
+import org.apache.pdfbox.contentstream.operator.text.ShowTextLineAndSpace;
+import org.apache.pdfbox.contentstream.operator.text.MoveText;
+import org.apache.pdfbox.contentstream.operator.text.MoveTextSetLeading;
+import org.apache.pdfbox.contentstream.operator.text.NextLine;
+import org.apache.pdfbox.contentstream.operator.text.SetCharSpacing;
+import org.apache.pdfbox.contentstream.operator.text.SetTextLeading;
+import org.apache.pdfbox.contentstream.operator.text.SetTextRenderingMode;
+import org.apache.pdfbox.contentstream.operator.text.SetTextRise;
+import org.apache.pdfbox.contentstream.operator.text.SetWordSpacing;
+import org.apache.pdfbox.contentstream.operator.text.ShowText;
+import org.apache.pdfbox.cos.COSDictionary;
+import org.apache.pdfbox.pdmodel.font.PDFontDescriptor;
+
+/**
+ * LEGACY text calculations which are known to be incorrect but are depended on by PDFTextStripper.
+ *
+ * This class exists only so that we don't break the code of users who have their own subclasses of
+ * PDFTextStripper. It replaces the mostly empty implementation of showGlyph() in PDFStreamEngine
+ * with a heuristic implementation which is backwards compatible.
+ *
+ * DO NOT USE THIS CODE UNLESS YOU ARE WORKING WITH PDFTextStripper.
+ * THIS CODE IS DELIBERATELY INCORRECT, USE PDFStreamEngine INSTEAD.
+ */
+@SuppressWarnings({"PMD", "checkstyle:all"})
+class LegacyPDFStreamEngine extends PDFStreamEngine
+{
+    private static final Log LOG = LogFactory.getLog(LegacyPDFStreamEngine.class);
+
+    private int pageRotation;
+    private PDRectangle pageSize;
+    private Matrix translateMatrix;
+    private final GlyphList glyphList;
+    private final Map<COSDictionary, Float> fontHeightMap = new WeakHashMap<COSDictionary, Float>();
+
+    /**
+     * Constructor.
+     */
+    LegacyPDFStreamEngine() throws IOException
+    {
+        addOperator(new BeginText());
+        addOperator(new Concatenate());
+        addOperator(new DrawObject()); // special text version
+        addOperator(new EndText());
+        addOperator(new SetGraphicsStateParameters());
+        addOperator(new Save());
+        addOperator(new Restore());
+        addOperator(new NextLine());
+        addOperator(new SetCharSpacing());
+        addOperator(new MoveText());
+        addOperator(new MoveTextSetLeading());
+        addOperator(new SetFontAndSize());
+        addOperator(new ShowText());
+        addOperator(new ShowTextAdjusted());
+        addOperator(new SetTextLeading());
+        addOperator(new SetMatrix());
+        addOperator(new SetTextRenderingMode());
+        addOperator(new SetTextRise());
+        addOperator(new SetWordSpacing());
+        addOperator(new SetTextHorizontalScaling());
+        addOperator(new ShowTextLine());
+        addOperator(new ShowTextLineAndSpace());
+
+        // load additional glyph list for Unicode mapping
+        String path = "/org/apache/pdfbox/resources/glyphlist/additional.txt";
+        InputStream input = GlyphList.class.getResourceAsStream(path);
+        glyphList = new GlyphList(GlyphList.getAdobeGlyphList(), input);
+    }
+
+    /**
+     * This will initialize and process the contents of the stream.
+     *
+     * @param page the page to process
+     * @throws java.io.IOException if there is an error accessing the stream.
+     */
+    @Override
+    public void processPage(PDPage page) throws IOException
+    {
+        this.pageRotation = page.getRotation();
+        this.pageSize = page.getCropBox();
+
+        if (pageSize.getLowerLeftX() == 0 && pageSize.getLowerLeftY() == 0)
+        {
+            translateMatrix = null;
+        }
+        else
+        {
+            // translation matrix for cropbox
+            translateMatrix = Matrix.getTranslateInstance(-pageSize.getLowerLeftX(), -pageSize.getLowerLeftY());
+        }
+        super.processPage(page);
+    }
+
+    /**
+     * Called when a glyph is to be processed. The heuristic calculations here were originally
+     * written by Ben Litchfield for PDFStreamEngine.
+     */
+    @Override
+    protected void showGlyph(Matrix textRenderingMatrix, PDFont font, int code,
+            String unicode,
+            Vector displacement)
+            throws IOException
+    {
+        //
+        // legacy calculations which were previously in PDFStreamEngine
+        //
+        //  DO NOT USE THIS CODE UNLESS YOU ARE WORKING WITH PDFTextStripper.
+        //  THIS CODE IS DELIBERATELY INCORRECT
+        //
+
+        PDGraphicsState state = getGraphicsState();
+        Matrix ctm = state.getCurrentTransformationMatrix();
+        float fontSize = state.getTextState().getFontSize();
+        float horizontalScaling = state.getTextState().getHorizontalScaling() / 100f;
+        Matrix textMatrix = getTextMatrix();
+
+        float displacementX = displacement.getX();
+        // the sorting algorithm is based on the width of the character. As the displacement
+        // for vertical characters doesn't provide any suitable value for it, we have to 
+        // calculate our own
+        if (font.isVertical())
+        {
+            displacementX = font.getWidth(code) / 1000;
+            // there may be an additional scaling factor for true type fonts
+            TrueTypeFont ttf = null;
+            if (font instanceof PDTrueTypeFont)
+            {
+                ttf = ((PDTrueTypeFont)font).getTrueTypeFont();
+            }
+            else if (font instanceof PDType0Font)
+            {
+                PDCIDFont cidFont = ((PDType0Font)font).getDescendantFont();
+                if (cidFont instanceof PDCIDFontType2)
+                {
+                    ttf = ((PDCIDFontType2)cidFont).getTrueTypeFont();
+                }
+            }
+            if (ttf != null && ttf.getUnitsPerEm() != 1000)
+            {
+                displacementX *= 1000f / ttf.getUnitsPerEm();
+            }
+        }
+
+        //
+        // legacy calculations which were previously in PDFStreamEngine
+        //
+        //  DO NOT USE THIS CODE UNLESS YOU ARE WORKING WITH PDFTextStripper.
+        //  THIS CODE IS DELIBERATELY INCORRECT
+        //
+
+        // (modified) combined displacement, this is calculated *without* taking the character
+        // spacing and word spacing into account, due to legacy code in TextStripper
+        float tx = displacementX * fontSize * horizontalScaling;
+        float ty = displacement.getY() * fontSize;
+
+        // (modified) combined displacement matrix
+        Matrix td = Matrix.getTranslateInstance(tx, ty);
+
+        // (modified) text rendering matrix
+        Matrix nextTextRenderingMatrix = td.multiply(textMatrix).multiply(ctm); // text space -> device space
+        float nextX = nextTextRenderingMatrix.getTranslateX();
+        float nextY = nextTextRenderingMatrix.getTranslateY();
+
+        // (modified) width and height calculations
+        float dxDisplay = nextX - textRenderingMatrix.getTranslateX();
+        Float fontHeight = fontHeightMap.get(font.getCOSObject());
+        if (fontHeight == null)
+        {
+            fontHeight = computeFontHeight(font);
+            fontHeightMap.put(font.getCOSObject(), fontHeight);
+        }
+        float dyDisplay = fontHeight * textRenderingMatrix.getScalingFactorY();
+
+        //
+        // start of the original method
+        //
+
+        // Note on variable names. There are three different units being used in this code.
+        // Character sizes are given in glyph units, text locations are initially given in text
+        // units, and we want to save the data in display units. The variable names should end with
+        // Text or Disp to represent if the values are in text or disp units (no glyph units are
+        // saved).
+
+        float glyphSpaceToTextSpaceFactor = 1 / 1000f;
+        if (font instanceof PDType3Font)
+        {
+            glyphSpaceToTextSpaceFactor = font.getFontMatrix().getScaleX();
+        }
+
+        float spaceWidthText = 0;
+        try
+        {
+            // to avoid crash as described in PDFBOX-614, see what the space displacement should be
+            spaceWidthText = font.getSpaceWidth() * glyphSpaceToTextSpaceFactor;
+        }
+        catch (Throwable exception)
+        {
+            LOG.warn(exception, exception);
+        }
+
+        if (spaceWidthText == 0)
+        {
+            spaceWidthText = font.getAverageFontWidth() * glyphSpaceToTextSpaceFactor;
+            // the average space width appears to be higher than necessary so make it smaller
+            spaceWidthText *= .80f;
+        }
+        if (spaceWidthText == 0)
+        {
+            spaceWidthText = 1.0f; // if could not find font, use a generic value
+        }
+
+        // the space width has to be transformed into display units
+        float spaceWidthDisplay = spaceWidthText * textRenderingMatrix.getScalingFactorX();
+
+        // use our additional glyph list for Unicode mapping
+        String unicodeMapping = font.toUnicode(code, glyphList);
+
+        // when there is no Unicode mapping available, Acrobat simply coerces the character code
+        // into Unicode, so we do the same. Subclasses of PDFStreamEngine don't necessarily want
+        // this, which is why we leave it until this point in PDFTextStreamEngine.
+        if (unicodeMapping == null)
+        {
+            if (font instanceof PDSimpleFont)
+            {
+                char c = (char) code;
+                unicodeMapping = new String(new char[] { c });
+            }
+            else
+            {
+                // Acrobat doesn't seem to coerce composite font's character codes, instead it
+                // skips them. See the "allah2.pdf" TestTextStripper file.
+                return;
+            }
+        }
+
+        // adjust for cropbox if needed
+        Matrix translatedTextRenderingMatrix;
+        if (translateMatrix == null)
+        {
+            translatedTextRenderingMatrix = textRenderingMatrix;
+        }
+        else
+        {
+            translatedTextRenderingMatrix = Matrix.concatenate(translateMatrix, textRenderingMatrix);
+            nextX -= pageSize.getLowerLeftX();
+            nextY -= pageSize.getLowerLeftY();
+        }
+
+        processTextPosition(new TextPosition(pageRotation, pageSize.getWidth(),
+                pageSize.getHeight(), translatedTextRenderingMatrix, nextX, nextY,
+                Math.abs(dyDisplay), dxDisplay,
+                Math.abs(spaceWidthDisplay), unicodeMapping, new int[] { code }, font,
+                fontSize,
+                (int)(fontSize * textMatrix.getScalingFactorX())));
+    }
+
+    /**
+     * Compute the font height. Override this if you want to use own calculations.
+     * 
+     * @param font the font.
+     * @return the font height.
+     * 
+     * @throws IOException if there is an error while getting the font bounding box.
+     */
+    protected float computeFontHeight(PDFont font) throws IOException
+    {
+        BoundingBox bbox = font.getBoundingBox();
+        if (bbox.getLowerLeftY() < Short.MIN_VALUE)
+        {
+            // PDFBOX-2158 and PDFBOX-3130
+            // files by Salmat eSolutions / ClibPDF Library
+            bbox.setLowerLeftY(- (bbox.getLowerLeftY() + 65536));
+        }
+        // 1/2 the bbox is used as the height todo: why?
+        float glyphHeight = bbox.getHeight() / 2;
+
+        // sometimes the bbox has very high values, but CapHeight is OK
+        PDFontDescriptor fontDescriptor = font.getFontDescriptor();
+        if (fontDescriptor != null)
+        {
+            float capHeight = fontDescriptor.getCapHeight();
+            if (Float.compare(capHeight, 0) != 0 &&
+                    (capHeight < glyphHeight || Float.compare(glyphHeight, 0) == 0))
+            {
+                glyphHeight = capHeight;
+            }
+            // PDFBOX-3464, PDFBOX-4480, PDFBOX-4553:
+            // sometimes even CapHeight has very high value, but Ascent and Descent are ok
+            float ascent = fontDescriptor.getAscent();
+            float descent = fontDescriptor.getDescent();
+            if (capHeight > ascent && ascent > 0 && descent < 0 &&
+                    ((ascent - descent) / 2 < glyphHeight || Float.compare(glyphHeight, 0) == 0))
+            {
+                glyphHeight = (ascent - descent) / 2;
+            }
+        }
+
+        // transformPoint from glyph space -> text space
+        float height;
+        if (font instanceof PDType3Font)
+        {
+            height = font.getFontMatrix().transformPoint(0, glyphHeight).y;
+        }
+        else
+        {
+            height = glyphHeight / 1000;
+        }
+
+        return height;
+    }
+
+    /**
+     * A method provided as an event interface to allow a subclass to perform some specific
+     * functionality when text needs to be processed.
+     *
+     * @param text The text to be processed.
+     */
+    protected void processTextPosition(TextPosition text)
+    {
+        // subclasses can override to provide specific functionality
+    }
+}
@@ -36,7 +36,6 @@ import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.pdmodel.graphics.PDXObject;
 import org.apache.pdfbox.pdmodel.graphics.image.PDImageXObject;
-import org.apache.pdfbox.text.PDFTextStripper;
 import org.apache.pdfbox.text.TextPosition;
 import org.apache.pdfbox.util.Matrix;

@@ -125,11 +124,9 @@ public class PDFLinesTextStripper extends PDFTextStripper {

                    // The direction of vertical lines must always be from bottom to top for the table extraction algorithm.
                    if (pos.getY() > path_y) {
-                        graphicsPath.add(new Ruling(new Point2D.Float(path_x, path_y), new Point2D.Float((float) pos.getX(), (float) pos
-                                .getY())));
+                        graphicsPath.add(new Ruling(new Point2D.Float(path_x, path_y), new Point2D.Float((float) pos.getX(), (float) pos.getY())));
                    } else {
-                        graphicsPath.add(new Ruling(new Point2D.Float(path_x, (float) pos.getY()), new Point2D.Float((float) pos
-                                .getX(), path_y)));
+                        graphicsPath.add(new Ruling(new Point2D.Float(path_x, (float) pos.getY()), new Point2D.Float((float) pos.getX(), path_y)));
                    }

                    path_x = (float) pos.getX();
@@ -150,25 +147,19 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                    Point2D p2 = transformPosition(x + width, y + height);

                    // Horizontal lines
-                    graphicsPath.add(new Ruling(new Point2D.Float((float) p1.getX(), (float) p1.getY()), new Point2D.Float((float) p2
-                            .getX(), (float) p1.getY())));
-                    graphicsPath.add(new Ruling(new Point2D.Float((float) p1.getX(), (float) p2.getY()), new Point2D.Float((float) p2
-                            .getX(), (float) p2.getY())));
+                    graphicsPath.add(new Ruling(new Point2D.Float((float) p1.getX(), (float) p1.getY()), new Point2D.Float((float) p2.getX(), (float) p1.getY())));
+                    graphicsPath.add(new Ruling(new Point2D.Float((float) p1.getX(), (float) p2.getY()), new Point2D.Float((float) p2.getX(), (float) p2.getY())));

                    // Vertical lines, direction must always be from bottom to top for the table extraction algorithm.
                    if (p2.getY() > p1.getY()) {
-                        graphicsPath.add(new Ruling(new Point2D.Float((float) p2.getX(), (float) p1.getY()), new Point2D.Float((float) p2
-                                .getX(), (float) p2.getY())));
+                        graphicsPath.add(new Ruling(new Point2D.Float((float) p2.getX(), (float) p1.getY()), new Point2D.Float((float) p2.getX(), (float) p2.getY())));
                    } else {
-                        graphicsPath.add(new Ruling(new Point2D.Float((float) p2.getX(), (float) p2.getY()), new Point2D.Float((float) p2
-                                .getX(), (float) p1.getY())));
+                        graphicsPath.add(new Ruling(new Point2D.Float((float) p2.getX(), (float) p2.getY()), new Point2D.Float((float) p2.getX(), (float) p1.getY())));
                    }
                    if (p2.getY() > p1.getY()) {
-                        graphicsPath.add(new Ruling(new Point2D.Float((float) p1.getX(), (float) p1.getY()), new Point2D.Float((float) p1
-                                .getX(), (float) p2.getY())));
+                        graphicsPath.add(new Ruling(new Point2D.Float((float) p1.getX(), (float) p1.getY()), new Point2D.Float((float) p1.getX(), (float) p2.getY())));
                    } else {
-                        graphicsPath.add(new Ruling(new Point2D.Float((float) p1.getX(), (float) p2.getY()), new Point2D.Float((float) p1
-                                .getX(), (float) p1.getY())));
+                        graphicsPath.add(new Ruling(new Point2D.Float((float) p1.getX(), (float) p2.getY()), new Point2D.Float((float) p1.getX(), (float) p1.getY())));
                    }
                }
                break;
@@ -192,9 +183,9 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                graphicsPath.clear();
                break;

-            case OperatorName.DRAW_OBJECT:
-                processImageOperation(arguments);
-                break;
+//            case OperatorName.DRAW_OBJECT:
+//                processImageOperation(arguments);
+//                break;

        }

@@ -211,8 +202,7 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                PDImageXObject image = (PDImageXObject) xobject;
                Matrix ctmNew = getGraphicsState().getCurrentTransformationMatrix();

-                Rectangle2D rect = new Rectangle2D.Float(ctmNew.getTranslateX(), ctmNew.getTranslateY(), ctmNew.getScaleX(), ctmNew
-                        .getScaleY());
+                Rectangle2D rect = new Rectangle2D.Float(ctmNew.getTranslateX(), ctmNew.getTranslateY(), ctmNew.getScaleX(), ctmNew.getScaleY());

                // Memory Hack - sofReference kills me
                FieldUtils.writeField(image, "cachedImageSubsampling", -1, true);
@@ -295,9 +285,7 @@ public class PDFLinesTextStripper extends PDFTextStripper {

            if (i == 0 && (textPositions.get(i).getUnicode().equals(" ") || textPositions.get(i)
                    .getUnicode()
-                    .equals("\u00A0") || textPositions.get(i)
-                    .getUnicode()
-                    .equals("\t"))) {
+                    .equals("\u00A0") || textPositions.get(i).getUnicode().equals("\t"))) {
                startIndex++;
                continue;
            }
@@ -307,7 +295,9 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                List<TextPosition> sublist = textPositions.subList(startIndex, i);
                if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0)
                        .getUnicode()
-                        .equals(" ") || sublist.get(0).getUnicode().equals("\u00A0") || sublist.get(0).getUnicode().equals("\t")))) {
+                        .equals(" ") || sublist.get(0).getUnicode().equals("\u00A0") || sublist.get(0)
+                        .getUnicode()
+                        .equals("\t")))) {
                    textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
                }
                startIndex = i;
@@ -318,7 +308,9 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                List<TextPosition> sublist = textPositions.subList(startIndex, i);
                if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0)
                        .getUnicode()
-                        .equals(" ") || sublist.get(0).getUnicode().equals("\u00A0") || sublist.get(0).getUnicode().equals("\t")))) {
+                        .equals(" ") || sublist.get(0).getUnicode().equals("\u00A0") || sublist.get(0)
+                        .getUnicode()
+                        .equals("\t")))) {
                    textPositionSequences.add(new TextPositionSequence(sublist, pageNumber));
                }
                startIndex = i;
@@ -332,7 +324,9 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                List<TextPosition> sublist = textPositions.subList(startIndex, i);
                if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0)
                        .getUnicode()
-                        .equals(" ") || sublist.get(0).getUnicode().equals("\u00A0") || sublist.get(0).getUnicode().equals("\t")))) {
+                        .equals(" ") || sublist.get(0).getUnicode().equals("\u00A0") || sublist.get(0)
+                        .getUnicode()
+                        .equals("\t")))) {

                    // Remove false sequence ends (whitespaces)
                    if (previous != null && sublist.get(0).getYDirAdj() == previous.getYDirAdj() && sublist.get(0)
@@ -351,15 +345,15 @@ public class PDFLinesTextStripper extends PDFTextStripper {
        List<TextPosition> sublist = textPositions.subList(startIndex, textPositions.size());
        if (!sublist.isEmpty() && (sublist.get(sublist.size() - 1)
                .getUnicode()
-                .equals(" ") || sublist.get(sublist.size() - 1).getUnicode().equals("\u00A0") || sublist.get(sublist.size() - 1).getUnicode().equals("\t"))) {
+                .equals(" ") || sublist.get(sublist.size() - 1)
+                .getUnicode()
+                .equals("\u00A0") || sublist.get(sublist.size() - 1).getUnicode().equals("\t"))) {
            sublist = sublist.subList(0, sublist.size() - 1);
        }

        if (!(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0).getUnicode().equals(" ") || sublist.get(0)
                .getUnicode()
-                .equals("\u00A0") || sublist.get(0)
-                .getUnicode()
-                .equals("\t")))) {
+                .equals("\u00A0") || sublist.get(0).getUnicode().equals("\t")))) {
            if (previous != null && sublist.get(0).getYDirAdj() == previous.getYDirAdj() && sublist.get(0)
                    .getXDirAdj() - (previous.getXDirAdj() + previous.getWidthDirAdj()) < 0.01) {
                for (TextPosition t : sublist) {
@@ -1,22 +1,24 @@
 package com.iqser.red.service.redaction.v1.server.parsing.model;

-import static java.util.stream.Collectors.toSet;
-
-import com.fasterxml.jackson.annotation.JsonIgnore;
-import com.fasterxml.jackson.annotation.JsonIgnoreProperties;
-import com.iqser.red.service.redaction.v1.model.Point;
-import com.iqser.red.service.redaction.v1.model.Rectangle;
-import lombok.Data;
-import lombok.NoArgsConstructor;
-import org.apache.pdfbox.text.TextPosition;
-
 import java.util.ArrayList;
 import java.util.List;
 import java.util.stream.Collectors;

+import org.apache.pdfbox.text.TextPosition;
+
+import com.fasterxml.jackson.annotation.JsonIgnore;
+import com.fasterxml.jackson.annotation.JsonIgnoreProperties;
+import com.iqser.red.service.redaction.v1.model.Point;
+import com.iqser.red.service.redaction.v1.model.Rectangle;
+
+import lombok.Data;
+import lombok.NoArgsConstructor;
+import lombok.extern.slf4j.Slf4j;
+
+@Slf4j
@Data
@NoArgsConstructor
-@JsonIgnoreProperties({ "empty" })
+@JsonIgnoreProperties({"empty"})
 public class TextPositionSequence implements CharSequence {

    private int page;
@@ -25,12 +27,15 @@ public class TextPositionSequence implements CharSequence {
    private float x1;
    private float x2;

+
    public TextPositionSequence(int page) {
+
        this.page = page;
    }


    public static TextPositionSequence fromData(List<RedTextPosition> textPositions, int page) {
+
        var textPositionSequence = new TextPositionSequence();
        textPositionSequence.textPositions = textPositions;
        textPositionSequence.page = page;
@@ -46,9 +51,6 @@ public class TextPositionSequence implements CharSequence {
    }


-
-
-
    @Override
    public int length() {

@@ -131,6 +133,7 @@ public class TextPositionSequence implements CharSequence {
        }
    }

+
    @JsonIgnore
    public float getRotationAdjustedY() {

@@ -190,10 +193,8 @@ public class TextPositionSequence implements CharSequence {

    @JsonIgnore
    public String getFont() {
-        return textPositions.get(0).getFontName()
-                .toLowerCase()
-                .replaceAll(",bold", "")
-                .replaceAll(",italic", "");
+
+        return textPositions.get(0).getFontName().toLowerCase().replaceAll(",bold", "").replaceAll(",italic", "");
    }


@@ -214,27 +215,33 @@ public class TextPositionSequence implements CharSequence {

    }

+
    @JsonIgnore
    public float getFontSize() {

        return textPositions.get(0).getFontSizeInPt();
    }

+
    @JsonIgnore
    public float getSpaceWidth() {

        return textPositions.get(0).getWidthOfSpace();
    }

+
    @JsonIgnore
    public int getRotation() {

        return textPositions.get(0).getRotation();
    }

+
    @JsonIgnore
    public Rectangle getRectangle() {

+        log.debug("Page: '{}', Word: '{}', Rotation: '{}'", page, toString(), textPositions.get(0).getRotation());
+
        float height = getTextHeight();

        float posXInit = getX1();
@@ -246,36 +253,45 @@ public class TextPositionSequence implements CharSequence {
            posXEnd = textPositions.get(0).getYDirAdj() + 2;
            posYInit = getY1();
            posYEnd = textPositions.get(textPositions.size() - 1).getXDirAdj() - height + 4;
-        } else if (textPositions.get(0).getRotation() == 270) {
+
+        } else if (textPositions.get(0).getRotation() == 270 && textPositions.get(0).getDir() == 270.0f) {
            posYInit = textPositions.get(0).getPageHeight() - getX1();
            posYEnd = textPositions.get(0).getPageHeight() - getX2() - textPositions.get(0)
                    .getWidth() - textPositions.get(textPositions.size() - 1).getWidth() - 1;
            posXInit = textPositions.get(0).getPageWidth() - textPositions.get(0).getYDirAdj() - 2;
            posXEnd = textPositions.get(0).getPageWidth() - textPositions.get(textPositions.size() - 1)
                    .getYDirAdj() + height;
-        } else if(textPositions.get(0).getRotation() == 0 && textPositions.get(0).getDir() == 270f) {
+
+        } else if (textPositions.get(0).getRotation() == 270 && textPositions.get(0).getDir() == 0.0f) {
+            posYInit = textPositions.get(0).getPageHeight() - textPositions.get(0).getYDirAdj() - 2;
+            posYEnd = posYInit + 1;
+            posXInit = textPositions.get(0).getXDirAdj();
+            posXEnd = textPositions.get(textPositions.size() - 1).getXDirAdj() + textPositions.get(textPositions.size() - 1).getWidthDirAdj() + 0.1f;
+
+        } else if (textPositions.get(0).getRotation() == 0 && textPositions.get(0).getDir() == 270f) {
            posYInit = textPositions.get(0).getPageHeight() - getX1();
            posYEnd = textPositions.get(0).getPageHeight() - getX2() - textPositions.get(0)
                    .getWidthDirAdj() - textPositions.get(textPositions.size() - 1).getWidthDirAdj() - 3;
            posXInit = textPositions.get(0).getPageWidth() - textPositions.get(0).getYDirAdj() - 2;
            posXEnd = textPositions.get(0).getPageWidth() - textPositions.get(textPositions.size() - 1)
                    .getYDirAdj() + height;
-        } else if(textPositions.get(0).getRotation() == 90 && textPositions.get(0).getDir() == 0.0f){
-            posXInit =  textPositions.get(textPositions.size() - 1)
+
+        } else if (textPositions.get(0).getRotation() == 90 && textPositions.get(0).getDir() == 0.0f) {
+            posXInit = textPositions.get(textPositions.size() - 1)
                    .getXDirAdj() + textPositions.get(textPositions.size() - 1).getHeightDir();
            posXEnd = textPositions.get(0).getXDirAdj();
            posYInit = textPositions.get(0).getPageHeight() - textPositions.get(0).getYDirAdj() - 2;
            posYEnd = textPositions.get(0).getPageHeight() - textPositions.get(textPositions.size() - 1)
                    .getYDirAdj() + 2;
-        } else if(textPositions.get(0).getRotation() == 0 && textPositions.get(0).getDir() == 90f){
-            posYInit =  getX1();
-            posYEnd =   getX2() + textPositions.get(0)
-                    .getWidthDirAdj() - textPositions.get(textPositions.size() - 1).getWidthDirAdj() - 3;
-            posXInit =  textPositions.get(0).getYDirAdj() + 2;
-            posXEnd =  textPositions.get(textPositions.size() - 1)
-                    .getYDirAdj() - height;
-        }
-        else {
+
+        } else if (textPositions.get(0).getRotation() == 0 && textPositions.get(0).getDir() == 90f) {
+            posYInit = getX1();
+            posYEnd = getX2() + textPositions.get(0).getWidthDirAdj() - textPositions.get(textPositions.size() - 1)
+                    .getWidthDirAdj() - 3;
+            posXInit = textPositions.get(0).getYDirAdj() + 2;
+            posXEnd = textPositions.get(textPositions.size() - 1).getYDirAdj() - height;
+
+        } else {
            posXEnd = textPositions.get(textPositions.size() - 1)
                    .getXDirAdj() + textPositions.get(textPositions.size() - 1).getWidthDirAdj() + 1;
            posYInit = textPositions.get(0).getPageHeight() - textPositions.get(0).getYDirAdj() - 2;
@@ -283,7 +299,9 @@ public class TextPositionSequence implements CharSequence {
                    .getYDirAdj() + 2;
        }

-        return new Rectangle(new Point(posXInit, posYInit), posXEnd - posXInit, posYEnd - posYInit + height, page);
+        var rectangle = new Rectangle(new Point(posXInit, posYInit), posXEnd - posXInit, posYEnd - posYInit + height, page);
+        log.debug("Rectangle: {}", rectangle);
+        return rectangle;
    }

 }
@@ -7,6 +7,7 @@ import com.iqser.red.service.redaction.v1.model.AnalyzeResult;
 import com.iqser.red.service.redaction.v1.model.StructureAnalyzeRequest;
 import com.iqser.red.service.redaction.v1.server.client.FileStatusProcessingUpdateClient;
 import com.iqser.red.service.redaction.v1.server.redaction.service.AnalyzeService;
+import com.iqser.red.service.redaction.v1.server.redaction.service.ManualRedactionSurroundingTextService;
 import com.iqser.red.service.redaction.v1.server.redaction.service.NerAnalyserService;

 import lombok.RequiredArgsConstructor;
@@ -28,6 +29,7 @@ public class RedactionMessageReceiver {
    private final AnalyzeService analyzeService;
    private final FileStatusProcessingUpdateClient fileStatusProcessingUpdateClient;
    private final NerAnalyserService nerAnalyserService;
+    private final ManualRedactionSurroundingTextService manualRedactionSurroundingTextService;


    @RabbitHandler
@@ -36,22 +38,37 @@ public class RedactionMessageReceiver {

        var analyzeRequest = objectMapper.readValue(in, AnalyzeRequest.class);
        log.info("Processing analyze request for file: {}", analyzeRequest.getFileId());
-        AnalyzeResult result;
-        if (analyzeRequest.isReanalyseOnlyIfPossible()) {
-            result = analyzeService.reanalyze(analyzeRequest);
-            log.info("Successfully reanalyzed dossier {} file {} took: {}", analyzeRequest.getDossierId(), analyzeRequest.getFileId(), result.getDuration());
-        } else {
-            // TODO Seperate stucture analysis by other queue
-            analyzeService.analyzeDocumentStructure(new StructureAnalyzeRequest(analyzeRequest.getDossierId(), analyzeRequest.getFileId()));
+        AnalyzeResult result = null;

-            // TODO NerEntities should be computed and stored in entity-recognition-service, should be triggered by a seperate queue after structure analysis
-            nerAnalyserService.computeNerEntities(analyzeRequest.getDossierId(), analyzeRequest.getFileId());
+        switch (analyzeRequest.getMessageType()){

-            result = analyzeService.analyze(analyzeRequest);
-            log.info("Successfully analyzed dossier {} file {} took: {}", analyzeRequest.getDossierId(), analyzeRequest.getFileId(), result
-                    .getDuration());
+            case REANALYSE:
+                result = analyzeService.reanalyze(analyzeRequest);
+                log.info("Successfully reanalyzed dossier {} file {} took: {}", analyzeRequest.getDossierId(), analyzeRequest.getFileId(), result.getDuration());
+                break;
+            case FULL_ANALYSE:
+                // TODO Seperate stucture analysis by other queue
+                analyzeService.analyzeDocumentStructure(new StructureAnalyzeRequest(analyzeRequest.getDossierId(), analyzeRequest.getFileId()));
+
+                // TODO NerEntities should be computed and stored in entity-recognition-service, should be triggered by a seperate queue after structure analysis
+                nerAnalyserService.computeNerEntities(analyzeRequest.getDossierId(), analyzeRequest.getFileId());
+
+                result = analyzeService.analyze(analyzeRequest);
+                log.info("Successfully analyzed dossier {} file {} took: {}", analyzeRequest.getDossierId(), analyzeRequest.getFileId(), result
+                        .getDuration());
+
+                break;
+            case SURROUNDING_TEXT:
+                var manualRedactions = manualRedactionSurroundingTextService.addSurroundingText(analyzeRequest.getDossierId(), analyzeRequest.getFileId(), analyzeRequest.getManualRedactions());
+                result = AnalyzeResult.builder()
+                        .dossierId(analyzeRequest.getDossierId())
+                        .fileId(analyzeRequest.getFileId())
+                        .manualRedactions(manualRedactions)
+                        .build();
+                break;
        }

+        result.setMessageType(analyzeRequest.getMessageType());
        fileStatusProcessingUpdateClient.analysisSuccessful(analyzeRequest.getDossierId(), analyzeRequest.getFileId(), result);
    }

@@ -42,6 +42,8 @@ public class Entity implements ReasonHolder {

    private boolean isDossierDictionaryEntry;

+    private boolean ignored;
+
    private Set<Engine> engines = new HashSet<>();

    private Set<Entity> references = new HashSet<>();
@@ -20,6 +20,7 @@ public class Image implements ReasonHolder {
    private int sectionNumber;
    private String section;
    private int page;
+    private boolean ignored;
    private boolean hasTransparency;

 }
@@ -16,6 +16,7 @@ public class PdfImage {
    private BufferedImage image;
    @NonNull
    private RedRectangle2D position;
+    @NonNull
    private ImageType imageType;
    private boolean isAppendedToParagraph;
    private boolean hasTransparency;
@@ -156,14 +156,14 @@ public class Section {
    @WhenCondition
    public boolean matchesType(@Argument(ArgumentType.TYPE) String type) {

-        return entities.stream().anyMatch(entity -> entity.getType().equals(type));
+        return entities.stream().anyMatch(entity -> !entity.isIgnored() && entity.getType().equals(type));
    }


    @WhenCondition
    public boolean matchesImageType(@Argument(ArgumentType.TYPE) String type) {

-        return images.stream().anyMatch(image -> image.getType().equals(type));
+        return images.stream().anyMatch(image -> !image.isIgnored() && image.getType().equals(type));
    }


@@ -175,27 +175,27 @@ public class Section {


    @ThenAction
-    public void expandByRegEx(@Argument(ArgumentType.TYPE) String type, @Argument(ArgumentType.REGEX) String pattern,
+    public void expandByRegEx(@Argument(ArgumentType.TYPE) String type, @Argument(ArgumentType.REGEX) String suffixPattern,
                              @Argument(ArgumentType.BOOLEAN) boolean patternCaseInsensitive,
                              @Argument(ArgumentType.INTEGER) int group) {

-        expandByRegEx(type, pattern, patternCaseInsensitive, group, null);
+        expandByRegEx(type, suffixPattern, patternCaseInsensitive, group, null);
    }


    @ThenAction
-    public void expandByRegEx(@Argument(ArgumentType.TYPE) String type, @Argument(ArgumentType.REGEX) String pattern,
+    public void expandByRegEx(@Argument(ArgumentType.TYPE) String type, @Argument(ArgumentType.REGEX) String suffixPattern,
                              @Argument(ArgumentType.BOOLEAN) boolean patternCaseInsensitive,
                              @Argument(ArgumentType.INTEGER) int group,
-                              @Argument(ArgumentType.REGEX) String withoutPattern) {
+                              @Argument(ArgumentType.REGEX) String valuePattern) {

-        Pattern compiledWithoutPattern = null;
+        Pattern compiledValuePattern = null;

-        if (withoutPattern != null) {
-            compiledWithoutPattern = Patterns.getCompiledPattern(withoutPattern, patternCaseInsensitive);
+        if (valuePattern != null) {
+            compiledValuePattern = Patterns.getCompiledPattern(valuePattern, patternCaseInsensitive);
        }

-        Pattern compiledPattern = Patterns.getCompiledPattern(pattern, patternCaseInsensitive);
+        Pattern compiledPattern = Patterns.getCompiledPattern(suffixPattern, patternCaseInsensitive);

        Set<Entity> expanded = new HashSet<>();
        for (Entity entity : entities) {
@@ -204,9 +204,9 @@ public class Section {
                continue;
            }

-            if (withoutPattern != null) {
-                Matcher matcherWithout = compiledWithoutPattern.matcher(entity.getWord());
-                if (matcherWithout.find()) {
+            if (valuePattern != null) {
+                Matcher valueMatcher = compiledValuePattern.matcher(entity.getWord());
+                if (!valueMatcher.matches()) {
                    continue;
                }
            }
@@ -0,0 +1,14 @@
+package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+
+import java.util.HashMap;
+import java.util.Map;
+
+import lombok.Data;
+
+@Data
+public class Classification {
+
+    private Map<String, Float> probabilities = new HashMap<>();
+    private String label;
+
+}
@@ -0,0 +1,10 @@
+package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+
+import lombok.Data;
+
+@Data
+public class FilterGeometry {
+
+    private ImageSize imageSize;
+    private ImageFormat imageFormat;
+}
@@ -0,0 +1,11 @@
+package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+
+import lombok.Data;
+
+@Data
+public class Filters {
+
+    private FilterGeometry geometry;
+    private Probability probability;
+    private boolean allPassed;
+}
@@ -0,0 +1,9 @@
+package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+
+import lombok.Data;
+
+@Data
+public class Geometry {
+    private float width;
+    private float height;
+}
@@ -0,0 +1,12 @@
+package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+
+import lombok.Data;
+
+@Data
+public class ImageFormat {
+
+    private float quotient;
+    private boolean tooTall;
+    private boolean tooWide;
+
+}
@@ -0,0 +1,12 @@
+package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+
+import lombok.Data;
+
+@Data
+public class ImageMetadata {
+
+    private Classification classification;
+    private Position position;
+    private Geometry geometry;
+    private Filters filters;
+}
@@ -0,0 +1,15 @@
+package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import lombok.Data;
+
+@Data
+public class ImageServiceResponse {
+
+    private String dossierId;
+    private String fileId;
+    private List<ImageMetadata> imageMetadata = new ArrayList<>();
+
+}
@@ -0,0 +1,12 @@
+package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+
+import lombok.Data;
+
+@Data
+public class ImageSize {
+
+    private float quotient;
+    private boolean tooLarge;
+    private boolean tooSmall;
+
+}
@@ -0,0 +1,12 @@
+package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+
+import lombok.Data;
+
+@Data
+public class Position {
+    private float x1;
+    private float x2;
+    private float y1;
+    private float y2;
+    private int pageNumber;
+}
@@ -0,0 +1,10 @@
+package com.iqser.red.service.redaction.v1.server.redaction.model.image;
+
+import lombok.Data;
+
+@Data
+public class Probability {
+
+    private boolean unconfident;
+
+}
@@ -26,8 +26,10 @@ import com.iqser.red.service.redaction.v1.server.redaction.model.DictionaryIncre
 import com.iqser.red.service.redaction.v1.server.redaction.model.DictionaryVersion;
 import com.iqser.red.service.redaction.v1.server.redaction.model.Image;
 import com.iqser.red.service.redaction.v1.server.redaction.model.PageEntities;
+import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
 import com.iqser.red.service.redaction.v1.server.redaction.model.RedRectangle2D;
 import com.iqser.red.service.redaction.v1.server.redaction.utils.EntitySearchUtils;
+import com.iqser.red.service.redaction.v1.server.segmentation.ImageService;
 import com.iqser.red.service.redaction.v1.server.segmentation.PdfSegmentationService;
 import com.iqser.red.service.redaction.v1.server.settings.RedactionServiceSettings;
 import com.iqser.red.service.redaction.v1.server.storage.RedactionStorageService;
@@ -53,6 +55,7 @@ public class AnalyzeService {
    private final SectionTextBuilderService sectionTextBuilderService;
    private final SectionGridCreatorService sectionGridCreatorService;
    private final NerAnalyserService nerAnalyserService;
+    private final ImageService imageService;


    public void analyzeDocumentStructure(StructureAnalyzeRequest analyzeRequest) {
@@ -63,9 +66,15 @@ public class AnalyzeService {
        Document classifiedDoc;

        try {
+
            var storedObjectStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(analyzeRequest
                    .getDossierId(), analyzeRequest.getFileId(), FileType.ORIGIN));
-            classifiedDoc = pdfSegmentationService.parseDocument(storedObjectStream);
+
+            Map<Integer, List<PdfImage>> pdfImages = null;
+            if(redactionServiceSettings.isEnableImageClassification()) {
+                pdfImages = imageService.convertImages(analyzeRequest.getDossierId(), analyzeRequest.getFileId());
+            }
+            classifiedDoc = pdfSegmentationService.parseDocument(storedObjectStream, pdfImages);
            pageCount = classifiedDoc.getPages().size();
        } catch (Exception e) {
            throw new RedactionException(e);
@@ -122,7 +122,7 @@ public class DictionaryService {

        dictionary.getDictionaryModels().forEach(dm -> {
            if (dm.isRecommendation() && !dm.getLocalEntries().isEmpty()) {
-                dictionaryClient.addEntries(IdBuilder.getTypeId(dm.getType(), dossierTemplateId, dm.isDossierDictionary() ? dossierId : null), new ArrayList<>(dm.getLocalEntries()), false);
+                dictionaryClient.addEntries(IdBuilder.getTypeId(dm.getType(), dossierTemplateId, dm.isDossierDictionary() ? dossierId : null), new ArrayList<>(dm.getLocalEntries()), false, true);
                long externalVersion = dictionaryClient.getVersion(dossierTemplateId);
                if (externalVersion == dictionary.getVersion().getDossierTemplateVersion() + 1) {
                    dictionary.getVersion().setDossierTemplateVersion(externalVersion);
@@ -10,7 +10,9 @@ import java.util.stream.Collectors;
 import java.util.stream.Stream;

 import com.iqser.red.service.persistence.service.v1.api.model.annotations.AnnotationStatus;
+import com.iqser.red.service.persistence.service.v1.api.model.annotations.IdRemoval;
 import com.iqser.red.service.persistence.service.v1.api.model.annotations.ManualImageRecategorization;
+
 import org.apache.commons.codec.binary.Base64;
 import org.apache.commons.lang3.StringUtils;
 import org.kie.api.runtime.KieContainer;
@@ -82,19 +84,42 @@ public class EntityRedactionService {
                surroundingWordsService.addSurroundingText(entities, reanalysisSection.getSearchableText(), dictionary);
            }

-            if (!local && reanalysisSection.getImages() != null && !reanalysisSection.getImages()
-                    .isEmpty() && analyzeRequest.getManualRedactions() != null && analyzeRequest.getManualRedactions()
-                    .getImageRecategorization() != null) {
-                for (Image image : reanalysisSection.getImages()) {
-                    String imageId = IdBuilder.buildId(image.getPosition(), image.getPage());
-                    for (ManualImageRecategorization imageRecategorization : analyzeRequest.getManualRedactions()
-                            .getImageRecategorization()) {
-                        if (imageRecategorization.getStatus().equals(AnnotationStatus.APPROVED) && imageRecategorization.getAnnotationId()
-                                .equals(imageId)) {
-                            image.setType(imageRecategorization.getType());
+            if (!local && analyzeRequest.getManualRedactions() != null) {
+
+                var approvedForceRedactions = analyzeRequest.getManualRedactions().getForceRedactions().stream()
+                        .filter(fr -> fr.getStatus() == AnnotationStatus.APPROVED)
+                        .filter(fr -> fr.getRequestDate() != null)
+                        .collect(Collectors.toList());
+                // only approved id removals, that haven't been forced back afterwards
+                var idsToRemove = analyzeRequest.getManualRedactions().getIdsToRemove().stream()
+                        .filter(idr -> idr.getStatus() == AnnotationStatus.APPROVED)
+                        .filter(idr -> idr.getRequestDate() != null)
+                        .filter(idr -> approvedForceRedactions.stream().noneMatch(forceRedact -> forceRedact.getRequestDate().isAfter(idr.getRequestDate())))
+                        .map(IdRemoval::getAnnotationId).collect(Collectors.toSet());
+
+                if (reanalysisSection.getImages() != null && !reanalysisSection.getImages()
+                        .isEmpty() && analyzeRequest.getManualRedactions().getImageRecategorization() != null) {
+                    for (Image image : reanalysisSection.getImages()) {
+                        String imageId = IdBuilder.buildId(image.getPosition(), image.getPage());
+                        for (ManualImageRecategorization imageRecategorization : analyzeRequest.getManualRedactions()
+                                .getImageRecategorization()) {
+                            if (imageRecategorization.getStatus()
+                                    .equals(AnnotationStatus.APPROVED) && imageRecategorization.getAnnotationId()
+                                    .equals(imageId)) {
+                                image.setType(imageRecategorization.getType());
+                            }
+                        }
+                        if (idsToRemove.contains(imageId)) {
+                            image.setIgnored(true);
                        }
                    }
                }
+
+                entities.forEach(entity -> entity.getPositionSequences().forEach(ps -> {
+                    if (idsToRemove.contains(ps.getId())) {
+                        entity.setIgnored(true);
+                    }
+                }));
            }

            sectionSearchableTextPairs.add(new SectionSearchableTextPair(Section.builder()
@@ -149,7 +174,8 @@ public class EntityRedactionService {
                        .add(new Entity(entity.getWord(), entity.getType(), entity.isRedaction(), entity.getRedactionReason(), entry
                                .getValue(), entity.getHeadline(), entity.getMatchedRule(), entity.getSectionNumber(), entity
                                .getLegalBasis(), entity.isDictionaryEntry(), entity.getTextBefore(), entity.getTextAfter(), entity
-                                .getStart(), entity.getEnd(), entity.isDossierDictionaryEntry(), entity.getEngines(), entity.getReferences()));
+                                .getStart(), entity.getEnd(), entity.isDossierDictionaryEntry(), entity.getEngines(), entity
+                                .getReferences()));
            }
        }
        return entitiesPerPage;
@@ -1,71 +0,0 @@
-package com.iqser.red.service.redaction.v1.server.redaction.service;
-
-import com.iqser.red.service.redaction.v1.server.classification.model.Page;
-import com.iqser.red.service.redaction.v1.server.client.ImageClassificationClient;
-import com.iqser.red.service.redaction.v1.server.client.ImageClassificationResponse;
-import com.iqser.red.service.redaction.v1.server.client.MockMultipartFile;
-import com.iqser.red.service.redaction.v1.server.redaction.model.ImageType;
-import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
-import com.iqser.red.service.redaction.v1.server.settings.RedactionServiceSettings;
-import lombok.RequiredArgsConstructor;
-import lombok.extern.slf4j.Slf4j;
-import org.springframework.stereotype.Service;
-
-import javax.imageio.ImageIO;
-import java.io.ByteArrayOutputStream;
-
-@Slf4j
-@Service
-@RequiredArgsConstructor
-public class ImageClassificationService {
-
-    private final ImageClassificationClient imageClassificationClient;
-    private final RedactionServiceSettings settings;
-
-
-    public void classifyImages(Page page) {
-
-        page.getImages().forEach(image -> {
-
-            if (settings.isEnableImageClassification() && !isEntirePageImage(image, page)) {
-
-                long start = System.currentTimeMillis();
-                try (ByteArrayOutputStream baos = new ByteArrayOutputStream()) {
-                    ImageIO.write(image.getImage(), "png", baos);
-                    var mockFile = new MockMultipartFile("file", "Image.png", "image/png", baos.toByteArray());
-                    ImageClassificationResponse response = imageClassificationClient.classify(mockFile);
-                    image.setImageType(ImageType.valueOf(response.getCategory()));
-                } catch (Exception e) {
-                    log.error("Could not classify image", e);
-                    image.setImageType(ImageType.OTHER);
-                }
-                log.info("Image classification took: " + (System.currentTimeMillis() - start));
-            } else {
-                image.setImageType(ImageType.OTHER);
-            }
-
-            image.getImage().flush();
-            image.setImage(null);
-
-            if (image.getImageType().equals(ImageType.OTHER)) {
-                page.getTextBlocks().forEach(textblock -> {
-                    if (image.getPosition()
-                            .contains(textblock.getMinX(), textblock.getMinY(), textblock.getWidth(), textblock.getHeight())) {
-                        image.setImageType(ImageType.OCR);
-                    }
-                });
-            }
-        });
-
-    }
-
-    private boolean isEntirePageImage(PdfImage image, Page page){
-        double imageArea = image.getPosition().getHeight() * image.getPosition().getWidth();
-        if(imageArea / page.getCropBoxArea() >= settings.getMaxImageCropboxRatio()){
-            log.info("Skipping image classification because images is almost as large as the entire page");
-            return true;
-        }
-        return false;
-    }
-
-}
@@ -0,0 +1,141 @@
+package com.iqser.red.service.redaction.v1.server.redaction.service;
+
+import java.util.ArrayList;
+import java.util.List;
+import java.util.Set;
+
+import org.apache.commons.lang3.tuple.Pair;
+import org.springframework.stereotype.Service;
+
+import com.iqser.red.service.persistence.service.v1.api.model.annotations.ManualRedactionEntry;
+import com.iqser.red.service.persistence.service.v1.api.model.annotations.ManualRedactions;
+import com.iqser.red.service.persistence.service.v1.api.model.annotations.ManualResizeRedaction;
+import com.iqser.red.service.persistence.service.v1.api.model.annotations.Rectangle;
+import com.iqser.red.service.redaction.v1.model.Engine;
+import com.iqser.red.service.redaction.v1.model.SectionArea;
+import com.iqser.red.service.redaction.v1.server.classification.model.SectionText;
+import com.iqser.red.service.redaction.v1.server.classification.model.Text;
+import com.iqser.red.service.redaction.v1.server.parsing.model.TextPositionSequence;
+import com.iqser.red.service.redaction.v1.server.redaction.model.Entity;
+import com.iqser.red.service.redaction.v1.server.redaction.model.EntityPositionSequence;
+import com.iqser.red.service.redaction.v1.server.redaction.utils.EntitySearchUtils;
+import com.iqser.red.service.redaction.v1.server.storage.RedactionStorageService;
+
+import lombok.RequiredArgsConstructor;
+
+@Service
+@RequiredArgsConstructor
+public class ManualRedactionSurroundingTextService {
+
+    private final RedactionStorageService redactionStorageService;
+    private final SurroundingWordsService surroundingWordsService;
+
+
+    public ManualRedactions addSurroundingText(String dossierId, String fileId, ManualRedactions manualRedactions) {
+
+        Text text = redactionStorageService.getText(dossierId, fileId);
+        List<ManualRedactionEntry> processedAddRedactions = new ArrayList<>();
+        List<ManualResizeRedaction> processedResizeRedactions = new ArrayList<>();
+
+        for (SectionText sectionText : text.getSectionTexts()) {
+
+            if (manualRedactions.getEntriesToAdd().isEmpty() && manualRedactions.getResizeRedactions().isEmpty()) {
+                break;
+            }
+
+            for (SectionArea sectionArea : sectionText.getSectionAreas()) {
+
+                var addItty = manualRedactions.getEntriesToAdd().iterator();
+                while (addItty.hasNext()) {
+                    var manualAddRedaction = addItty.next();
+                    if (sectionContainsEntry(sectionArea, manualAddRedaction.getPositions())) {
+                        var surroundingText = findSurroundingText(sectionText, manualAddRedaction.getValue(), manualAddRedaction
+                                .getPositions());
+                        manualAddRedaction.setTextBefore(surroundingText.getLeft());
+                        manualAddRedaction.setTextAfter(surroundingText.getRight());
+                        processedAddRedactions.add(manualAddRedaction);
+                        addItty.remove();
+                    }
+                }
+
+                var resizeItty = manualRedactions.getResizeRedactions().iterator();
+                while (resizeItty.hasNext()) {
+                    var manualResizeRedaction = resizeItty.next();
+                    if (sectionContainsEntry(sectionArea, manualResizeRedaction.getPositions())) {
+                        var surroundingText = findSurroundingText(sectionText, manualResizeRedaction.getValue(), manualResizeRedaction
+                                .getPositions());
+                        manualResizeRedaction.setTextBefore(surroundingText.getLeft());
+                        manualResizeRedaction.setTextAfter(surroundingText.getRight());
+                        processedResizeRedactions.add(manualResizeRedaction);
+                        resizeItty.remove();
+                    }
+                }
+            }
+        }
+
+        manualRedactions.getEntriesToAdd().addAll(processedAddRedactions);
+        manualRedactions.getResizeRedactions().addAll(processedResizeRedactions);
+        return manualRedactions;
+    }
+
+
+    private Pair<String, String> findSurroundingText(SectionText sectionText, String value,
+                                                     List<Rectangle> toFindPositions) {
+
+        Set<Entity> entities = EntitySearchUtils.find(sectionText.getText(), Set.of(value), "dummy", sectionText.getHeadline(), sectionText
+                .getSectionNumber(), false, false, Engine.DICTIONARY);
+        Set<Entity> entitiesWithPositions = EntitySearchUtils.clearAndFindPositions(entities, sectionText.getSearchableText(), null);
+
+        Entity correctEntity = getEntityOnCorrectPosition(entitiesWithPositions, toFindPositions);
+
+        if (sectionText.getCellStarts() != null && !sectionText.getCellStarts().isEmpty()) {
+            surroundingWordsService.addSurroundingText(Set.of(correctEntity), sectionText.getSearchableText(), null, sectionText
+                    .getCellStarts());
+        } else {
+            surroundingWordsService.addSurroundingText(Set.of(correctEntity), sectionText.getSearchableText(), null);
+        }
+
+        return Pair.of(correctEntity.getTextBefore(), correctEntity.getTextAfter());
+    }
+
+
+    private boolean sectionContainsEntry(SectionArea sectionArea, List<Rectangle> positions) {
+
+        for (Rectangle position : positions) {
+            if (sectionArea.contains(position)) {
+                return true;
+            }
+        }
+        return false;
+    }
+
+
+    private Entity getEntityOnCorrectPosition(Set<Entity> entitiesWithPositions, List<Rectangle> toFindPositions) {
+
+        for (Entity entityWithPos : entitiesWithPositions) {
+            for (EntityPositionSequence entityPositionSequence : entityWithPos.getPositionSequences()) {
+                for (TextPositionSequence textPositionSequence : entityPositionSequence.getSequences()) {
+                    for (Rectangle manualRedactionRectangle : toFindPositions) {
+                        if (intersects(manualRedactionRectangle, textPositionSequence.getRectangle())) {
+                            return entityWithPos;
+                        }
+                    }
+                }
+            }
+        }
+
+        return null;
+    }
+
+
+    public boolean intersects(Rectangle manualPosition,
+                              com.iqser.red.service.redaction.v1.model.Rectangle textPositionRectangle) {
+
+        return textPositionRectangle.getTopLeft()
+                .getX() + textPositionRectangle.getWidth() > manualPosition.getTopLeftX() && textPositionRectangle.getTopLeft()
+                .getY() + textPositionRectangle.getHeight() > manualPosition.getTopLeftY() && textPositionRectangle.getTopLeft()
+                .getX() < manualPosition.getTopLeftX() + manualPosition.getWidth() && textPositionRectangle.getTopLeft()
+                .getY() < manualPosition.getTopLeftY() + manualPosition.getHeight();
+    }
+
+}
@@ -50,7 +50,13 @@ public class RedactionChangeLogService {
                        .equals(ChangeType.REMOVED))
                .collect(Collectors.toList());

-        Set<RedactionLogEntry> added = new HashSet<>(currentRedactionLog.getRedactionLogEntry());
+        Set<RedactionLogEntry> added = new HashSet<>(currentRedactionLog.getRedactionLogEntry().stream()
+                .filter(entry -> entry.getChanges().isEmpty() || !entry.getChanges()
+                        .get(entry.getChanges().size() - 1)
+                        .getType()
+                        .equals(ChangeType.REMOVED))
+                .collect(Collectors.toList()));
+
        added.removeAll(notRemovedPreviousEntries);

        Set<RedactionLogEntry> removed = new HashSet<>(notRemovedPreviousEntries);
@@ -140,14 +140,14 @@ public class RedactionLogCreatorService {
        } else {
            float x = textPositions.get(0).getXDirAdj();
            float y = textPositions.get(0).getYDirAdj();
-            float width = textPositions.get(0).getWidth();
+            float width = textPositions.get(0).getWidthDirAdj();
            float height = textPositions.get(0).getHeightDir();
            int startIndex = 0;

            for (int i = 1; i < textPositions.size(); i++) {
                float xDirAdj = textPositions.get(i).getXDirAdj();
                float yDirAdj = textPositions.get(i).getYDirAdj();
-                float widthDir = textPositions.get(i).getWidth();
+                float widthDir = textPositions.get(i).getWidthDirAdj();
                float heightDir = textPositions.get(i).getHeightDir();

                if (!(isCharInSameLine(y, yDirAdj, height, heightDir) && isCharClose(x, xDirAdj, width))) {
@@ -23,8 +23,9 @@ public class RedactionLogMergeService {
    private final SectionTextService sectionTextService;


-    public RedactionLog mergeRedactionLogData(RedactionLog redactionLog, SectionGrid sectionGrid, String dossierTemplateId,
-                                              ManualRedactions manualRedactions, Set<Integer> excludedPages) {
+    public RedactionLog mergeRedactionLogData(RedactionLog redactionLog, SectionGrid sectionGrid,
+                                              String dossierTemplateId, ManualRedactions manualRedactions,
+                                              Set<Integer> excludedPages) {

        log.info("Merging Redaction log with manual redactions");
        if (manualRedactions != null) {
@@ -56,7 +57,7 @@ public class RedactionLogMergeService {

        Set<String> processedIds = new HashSet<>();
        redactionLog.getRedactionLogEntry().removeIf(entry -> {
-            if(processedIds.contains(entry.getId())){
+            if (processedIds.contains(entry.getId())) {
                log.info("Duplicate annotation found with id {}", entry.getId());
                return true;
            }
@@ -120,12 +121,19 @@ public class RedactionLogMergeService {
                    redactionLogEntry.setType(imageRecategorization.getType());
                    redactionLogEntry.setHasBeenRecategorized(true);
                    redactionLogEntry.setSection("Image:" + redactionLogEntry.getType());
+
+                    if (dictionaryService.isHint(imageRecategorization.getType(), dossierTemplateId)) {
+                        redactionLogEntry.setRedacted(false);
+                        redactionLogEntry.setHint(true);
+                    } else {
+                        redactionLogEntry.setHint(false);
+                    }
+
                    manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", recategorized by manual override");
                } else if (imageRecategorization.getStatus().equals(AnnotationStatus.REQUESTED)) {
                    manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", requested to recategorize");
                    redactionLogEntry.setStatus(AnnotationStatus.REQUESTED);
-                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, false, redactionLogEntry
-                            .isRedacted(), false));
+                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, false, redactionLogEntry.isRedacted(), false));
                    redactionLogEntry.setRecategorizationType(imageRecategorization.getType());
                } else {
                    redactionLogEntry.setStatus(AnnotationStatus.DECLINED);
@@ -139,47 +147,53 @@ public class RedactionLogMergeService {

            if (mrw.getItem() instanceof IdRemoval) {
                var manualRemoval = (IdRemoval) mrw.getItem();
-                String manualOverrideReason = null;
-                if (manualRemoval.getStatus().equals(AnnotationStatus.APPROVED)) {
-                    redactionLogEntry.setRedacted(false);
-                    redactionLogEntry.setStatus(AnnotationStatus.APPROVED);
-                    manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", removed by manual override");
-                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, false, redactionLogEntry
-                            .isRedacted(), true));
-                    redactionLogEntry.setHasBeenRemovedByManualOverride(true);
-                } else if (manualRemoval.getStatus().equals(AnnotationStatus.REQUESTED)) {
-                    manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", requested to remove");
-                    redactionLogEntry.setStatus(AnnotationStatus.REQUESTED);
-                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, true, redactionLogEntry
-                            .isRedacted(), false));
+                if (manualRemoval.getStatus().equals(AnnotationStatus.APPROVED) && manualRemoval.isRemoveFromDictionary()) {
+                    log.info("Skipping merge for dictionary-modifying entry");
                } else {
-                    redactionLogEntry.setStatus(AnnotationStatus.DECLINED);
-                }
+                    String manualOverrideReason = null;
+                    if (manualRemoval.getStatus().equals(AnnotationStatus.APPROVED)) {
+                        redactionLogEntry.setRedacted(false);
+                        redactionLogEntry.setStatus(AnnotationStatus.APPROVED);
+                        manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", removed by manual override");
+                        redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, false, redactionLogEntry.isRedacted(), true));
+                        redactionLogEntry.setHint(false);
+                        redactionLogEntry.setHasBeenRemovedByManualOverride(true);
+                    } else if (manualRemoval.getStatus().equals(AnnotationStatus.REQUESTED)) {
+                        manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", requested to remove");
+                        redactionLogEntry.setStatus(AnnotationStatus.REQUESTED);
+                        redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, true, redactionLogEntry.isRedacted(), false));
+                    } else {
+                        redactionLogEntry.setStatus(AnnotationStatus.DECLINED);
+                    }

-                redactionLogEntry.setReason(manualOverrideReason);
-                redactionLogEntry.setManual(true);
-                redactionLogEntry.setManualRedactionUserId(manualRemoval.getUser());
-                redactionLogEntry.setManualRedactionType(ManualRedactionType.REMOVE);
-                redactionLogEntry.setDictionaryEntry(manualRemoval.isRemoveFromDictionary());
-                redactionLogEntry.setDossierDictionaryEntry(manualRemoval.isRemoveFromDictionary());
+                    redactionLogEntry.setReason(manualOverrideReason);
+                    redactionLogEntry.setManual(true);
+                    redactionLogEntry.setManualRedactionUserId(manualRemoval.getUser());
+                    redactionLogEntry.setManualRedactionType(ManualRedactionType.REMOVE);
+                    redactionLogEntry.setDictionaryEntry(manualRemoval.isRemoveFromDictionary());
+                    redactionLogEntry.setDossierDictionaryEntry(manualRemoval.isRemoveFromDictionary());
+                }
            }

            if (mrw.getItem() instanceof ManualForceRedaction) {
                var manualForceRedact = (ManualForceRedaction) mrw.getItem();
                String manualOverrideReason = null;
                if (manualForceRedact.getStatus().equals(AnnotationStatus.APPROVED)) {
-                    redactionLogEntry.setRedacted(true);
+                    // Forcing a skipped hint should result in a hint
+                    if (dictionaryService.isHint(redactionLogEntry.getType(), dossierTemplateId)) {
+                        redactionLogEntry.setHint(true);
+                    } else {
+                        redactionLogEntry.setRedacted(true);
+                    }
                    redactionLogEntry.setStatus(AnnotationStatus.APPROVED);
-                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, false, redactionLogEntry
-                            .isRedacted(), false));
+                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, false, redactionLogEntry.isRedacted(), false));
                    manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", forced by manual override");
                    redactionLogEntry.setHasBeenForced(true);
                    redactionLogEntry.setLegalBasis(manualForceRedact.getLegalBasis());
                } else if (manualForceRedact.getStatus().equals(AnnotationStatus.REQUESTED)) {
                    manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", requested to force redact");
                    redactionLogEntry.setStatus(AnnotationStatus.REQUESTED);
-                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, true, redactionLogEntry
-                            .isRedacted(), false));
+                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, true, redactionLogEntry.isRedacted(), false));
                    redactionLogEntry.setLegalBasis(manualForceRedact.getLegalBasis());
                } else {
                    redactionLogEntry.setStatus(AnnotationStatus.DECLINED);
@@ -210,8 +224,7 @@ public class RedactionLogMergeService {
                } else if (manualLegalBasisChange.getStatus().equals(AnnotationStatus.REQUESTED)) {
                    manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", legal basis change requested");
                    redactionLogEntry.setStatus(AnnotationStatus.REQUESTED);
-                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, true, redactionLogEntry
-                            .isRedacted(), false));
+                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, true, redactionLogEntry.isRedacted(), false));
                    redactionLogEntry.setLegalBasisChangeValue(manualLegalBasisChange.getLegalBasis());
                } else {
                    redactionLogEntry.setStatus(AnnotationStatus.DECLINED);
@@ -228,18 +241,20 @@ public class RedactionLogMergeService {
                String manualOverrideReason = null;
                if (manualResizeRedact.getStatus().equals(AnnotationStatus.APPROVED)) {
                    redactionLogEntry.setStatus(AnnotationStatus.APPROVED);
-                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, false, redactionLogEntry
-                            .isRedacted(), false));
+                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, false, redactionLogEntry.isRedacted(), false));
                    redactionLogEntry.setPositions(convertPositions(manualResizeRedact.getPositions()));
                    redactionLogEntry.setValue(manualResizeRedact.getValue());
                    redactionLogEntry.setHasBeenResized(true);
+                    redactionLogEntry.setTextBefore(manualResizeRedact.getTextBefore());
+                    redactionLogEntry.setTextAfter(manualResizeRedact.getTextAfter());
                    manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", resized by manual override");
                } else if (manualResizeRedact.getStatus().equals(AnnotationStatus.REQUESTED)) {
                    manualOverrideReason = mergeReasonIfNecessary(redactionLogEntry.getReason(), ", requested to resize redact");
                    redactionLogEntry.setStatus(AnnotationStatus.REQUESTED);
-                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, true, redactionLogEntry
-                            .isRedacted(), false));
+                    redactionLogEntry.setColor(getColor(redactionLogEntry.getType(), dossierTemplateId, true, redactionLogEntry.isRedacted(), false));
                    redactionLogEntry.setPositions(convertPositions(manualResizeRedact.getPositions()));
+                    redactionLogEntry.setTextBefore(manualResizeRedact.getTextBefore());
+                    redactionLogEntry.setTextAfter(manualResizeRedact.getTextAfter());
                } else {
                    redactionLogEntry.setStatus(AnnotationStatus.DECLINED);
                }
@@ -280,6 +295,8 @@ public class RedactionLogMergeService {
                RedactionLogEntry redactionLogEntry = createRedactionLogEntry(manualRedactionEntry, manualRedactionEntry.getAnnotationId(), dossierTemplateId);
                redactionLogEntry.setPositions(convertPositions(manualRedactionEntry.getPositions()));
                redactionLogEntry.setComments(comments.get(manualRedactionEntry.getAnnotationId()));
+                redactionLogEntry.setTextBefore(manualRedactionEntry.getTextBefore());
+                redactionLogEntry.setTextAfter(manualRedactionEntry.getTextAfter());

                sectionTextService.handleSectionText(sectionGrid, redactionLogEntry);

@@ -290,9 +307,13 @@ public class RedactionLogMergeService {
        return redactionLogEntries;
    }

-    private List<Rectangle> convertPositions(List<com.iqser.red.service.persistence.service.v1.api.model.annotations.Rectangle> positions) {
-        return positions.stream().map(pos -> new Rectangle(new Point(pos.getTopLeftX(), pos.getTopLeftY()), pos.getWidth(), pos
-                .getHeight(), pos.getPage())).collect(Collectors.toList());
+
+    private List<Rectangle> convertPositions(
+            List<com.iqser.red.service.persistence.service.v1.api.model.annotations.Rectangle> positions) {
+
+        return positions.stream()
+                .map(pos -> new Rectangle(new Point(pos.getTopLeftX(), pos.getTopLeftY()), pos.getWidth(), pos.getHeight(), pos.getPage()))
+                .collect(Collectors.toList());
    }


@@ -375,6 +396,5 @@ public class RedactionLogMergeService {

    }

-
 }

@@ -28,7 +28,7 @@ public class SurroundingWordsService {
        try {
            for (Entity entity : entities) {

-                if (dictionary.isHint(entity.getType())) {
+                if (dictionary != null && dictionary.isHint(entity.getType())) {
                    continue;
                }
                findSurroundingWords(entity, searchableText.toString(), entity.getStart(), entity.getEnd());
@@ -64,7 +64,7 @@ public class SurroundingWordsService {
                    String text = searchableString.substring(startOffset, endOffset);
                    for (Entity entity : entities) {

-                        if (dictionary.isHint(entity.getType())) {
+                        if (dictionary != null && dictionary.isHint(entity.getType())) {
                            continue;
                        }

@@ -104,7 +104,7 @@ public class EntitySearchUtils {
                    .collect(Collectors.toList());
            Entity firstEntity = orderedEntities.get(0);
            List<EntityPositionSequence> positionSequences = text.getSequences(firstEntity.getWord()
-                    .trim(), dictionary.isCaseInsensitiveDictionary(firstEntity.getType()), firstEntity.getTargetSequences());
+                    .trim(), dictionary == null ? true : dictionary.isCaseInsensitiveDictionary(firstEntity.getType()), firstEntity.getTargetSequences());

            for (int i = 0; i <= orderedEntities.size() - 1; i++) {
                try {
@@ -193,15 +193,35 @@ public class EntitySearchUtils {

        Set<Entity> result = new HashSet<>();
        if (existingEntities != null && foundEntities != null) {
-            for (Entity existingEntity : existingEntities) {
-                for (Entity foundEntity : foundEntities) {
-                    if (existingEntity.getEnd() < foundEntity.getStart() || foundEntity.getEnd() < existingEntity.getStart()) {
-                        result.add(foundEntity);
-                    }
+            for (Entity foundEntity : foundEntities) {
+
+                if (!overlaps(existingEntities, foundEntity)) {
+                    result.add(foundEntity);
                }
+
            }
        }
        return result;
    }

+
+    private boolean overlaps(Set<Entity> existingEntities, Entity found) {
+
+        for (Entity existing : existingEntities) {
+
+            if(existing.getStart().equals(found.getStart())){
+                continue;
+            }
+
+            for (int i = existing.getStart(); i <= existing.getEnd(); i++) {
+                for (int j = found.getStart(); j <= found.getEnd(); j++) {
+                    if (i == j) {
+                        return true;
+                    }
+                }
+            }
+        }
+        return false;
+    }
+
 }
@@ -1,165 +0,0 @@
-package com.iqser.red.service.redaction.v1.server.segmentation;
-
-import java.awt.Graphics;
-import java.awt.geom.Rectangle2D;
-import java.awt.image.BufferedImage;
-import java.util.ArrayList;
-import java.util.List;
-
-import org.springframework.stereotype.Service;
-
-import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
-
-import lombok.RequiredArgsConstructor;
-import lombok.extern.slf4j.Slf4j;
-
-@Slf4j
-@Service
-@RequiredArgsConstructor
-public class ImageMergeService {
-
-
-    public List<PdfImage> mergeImages(List<PdfImage> images, int rotation){
-
-        List<PdfImage> mergedList = processImages(images, rotation);
-
-        List<PdfImage> imagesInImage = new ArrayList<>();
-        for(PdfImage image: mergedList){
-            for (PdfImage inner: mergedList){
-                if(image != inner && image.getPosition().contains(inner.getPosition().getX(), inner.getPosition().getY(), inner.getPosition().getWidth(), inner.getPosition().getHeight())){
-                    imagesInImage.add(inner);
-                }
-            }
-        }
-        mergedList.removeAll(imagesInImage);
-
-        return mergedList;
-    }
-
-
-    //merge images, if they are separated during pdf import, return new list of Pdfimages
-    private List<PdfImage> processImages(List<PdfImage> imageList, int rotation) {
-        if (imageList.size() > 1) {
-            List<PdfImage> mergedList = new ArrayList<>();
-            int countElementsInList = 0;
-            boolean beginImage = true;
-
-            // a List of Boolean, true = candidate for merging, false = no merging
-            List<Boolean> candidatesList = getCandidatesList(imageList, rotation);
-
-            // loop through list, if there are candidates for merging (true), merge images and add it to mergedList
-            for (int i = 0; i < candidatesList.size(); i++) {
-                if (candidatesList.get(i)) {
-                    if (beginImage) {
-                        //begin of image, merge two parts of imageList
-                        PdfImage mergedImage = mergeTwoImages(imageList.get(i), imageList.get(i + 1), rotation);
-                        // image merge successful
-                        if (mergedImage != null) {
-                            mergedList.add(mergedImage);
-                            countElementsInList++;
-                        }
-                    } else {
-                        //middle of an image, merge current piece auf mergedList with image of imageList
-                        PdfImage mergedImage = mergeTwoImages(mergedList.get(countElementsInList - 1), imageList.get(i + 1), rotation);
-                        // image merge successful
-                        if (mergedImage != null) {
-                            mergedList.set(countElementsInList - 1, mergedImage);
-                        }
-                    }
-                    beginImage = false;
-                } else {
-                    // if the last candidate is false, then both images i and i+1 must be added
-                    if (i == candidatesList.size() - 1) {
-                        if (countElementsInList > 0 && mergedList.get(countElementsInList - 1) == imageList.get(i)) {
-                            mergedList.add(imageList.get(i + 1));
-                        } else {
-                            mergedList.add(imageList.get(i));
-                            mergedList.add(imageList.get(i + 1));
-                        }
-                    } else {
-                        //first image is not splitted, add i to resultlist
-                        if (beginImage) {
-                            mergedList.add(imageList.get(i));
-                            countElementsInList++;
-                        } else {
-                            // i is the end of an image, add begin of new image
-                            mergedList.add(imageList.get(i + 1));
-                            countElementsInList++;
-                            beginImage = false;
-                        }
-                    }
-                }
-            }
-            return mergedList;
-        } else {
-            return imageList;
-        }
-    }
-
-    private PdfImage mergeTwoImages(PdfImage image1, PdfImage image2, int rotation) {
-
-        // diese Angaben von getPosition scheinen nicht richtig zu sein, damit werden teile des Bildes abgeschnitten
-        double width = image1.getPosition().getWidth();
-        double width2 = image2.getPosition().getWidth();
-        double height1 = image1.getPosition().getHeight();
-        double height2 = image2.getPosition().getHeight();
-        // mit  den Werten, die unter Image gespeichert sind, funktioniert es
-        double img1height = image1.getImage().getHeight();
-        double img1width = image1.getImage().getWidth();
-        double img2height = image2.getImage().getHeight();
-
-        BufferedImage mergedImage = new BufferedImage((int) img1width, (int) (img1height + img2height), BufferedImage.TYPE_INT_RGB);
-        Graphics mergedImageGraphics = mergedImage.getGraphics();
-        try {
-            mergedImageGraphics.drawImage(image1.getImage(), 0, 0, null);
-            mergedImageGraphics.drawImage(image2.getImage(), 0, (int) (img1height), null);
-
-            // set Image, Position and type for merged Image
-            //set position for merged image with values of image1 and the height of both
-            Rectangle2D pos = new Rectangle2D.Float();
-            pos.setRect(image1.getPosition().getX(), image2.getPosition().getY(), rotation == 90 ? width + width2: width, rotation == 90 ? height1 : height1 + height2);
-            PdfImage newPdfImage = new PdfImage(mergedImage, pos, image1.getPage(), image1.isHasTransparency() || image2.isHasTransparency());
-            // Graphics need to be disposed
-
-            image1.getImage().flush();
-            image2.getImage().flush();
-
-            mergedImage.flush();
-            mergedImageGraphics.dispose();
-
-            return newPdfImage;
-        } catch (Exception e) {
-            // failed to merge image
-            log.error("Failed to merge image", e);
-            return null;
-        }
-
-
-    }
-
-    //make a list of true and false, if the image is a candidate for merging
-    private List<Boolean> getCandidatesList(List<PdfImage> imageList, int rotation) {
-        List<Boolean> candidatesList = new ArrayList<>();
-        for (int i = 0; i < imageList.size(); i++) {
-            if (i >= 1) {
-                candidatesList.add(isCandidateForMerging(imageList.get(i - 1), imageList.get(i), rotation));
-            }
-        }
-        return candidatesList;
-    }
-
-    // evaluate if two images are candidates for merging, depending on their coordinates, width and height
-    private boolean isCandidateForMerging(PdfImage image1, PdfImage image2, int rotation) {
-        double x1 = rotation == 90 ? image1.getPosition().getY() : image1.getPosition().getX();
-        double y1 = rotation == 90 ? image1.getPosition().getX() : image1.getPosition().getY();
-        double width1 = rotation == 90 ? image1.getPosition().getHeight() : image1.getPosition().getWidth();
-        double x2 = rotation == 90 ? image2.getPosition().getY() : image2.getPosition().getX();
-        double y2 = rotation == 90 ? image2.getPosition().getX() : image2.getPosition().getY();
-        double width2 = rotation == 90 ? image2.getPosition().getHeight() : image2.getPosition().getWidth();
-        double height2 = rotation == 90 ? image2.getPosition().getWidth() : image2.getPosition().getHeight();
-        //if the x-coordinates and widths of images are equal and the height is equal to difference between y-coordinates,
-        // then it is the same picture and has to be merged -> return true
-        return x1 == x2 && width1 == width2 && Math.ceil(height2) == Math.ceil(rotation == 90 ? y2 - y1 : y1 - y2) && width2 > (height2 / 6);
-    }
-
-}
@@ -0,0 +1,61 @@
+package com.iqser.red.service.redaction.v1.server.segmentation;
+
+import java.util.ArrayList;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Locale;
+import java.util.Map;
+
+import org.springframework.stereotype.Service;
+
+import com.fasterxml.jackson.databind.ObjectMapper;
+import com.iqser.red.service.persistence.service.v1.api.model.dossiertemplate.dossier.file.FileType;
+import com.iqser.red.service.redaction.v1.server.classification.model.Page;
+import com.iqser.red.service.redaction.v1.server.redaction.model.ImageType;
+import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
+import com.iqser.red.service.redaction.v1.server.redaction.model.RedRectangle2D;
+import com.iqser.red.service.redaction.v1.server.redaction.model.image.ImageServiceResponse;
+import com.iqser.red.service.redaction.v1.server.storage.RedactionStorageService;
+
+import lombok.RequiredArgsConstructor;
+import lombok.SneakyThrows;
+
+@Service
+@RequiredArgsConstructor
+public class ImageService {
+
+    private final ObjectMapper objectMapper;
+    private final RedactionStorageService redactionStorageService;
+
+    @SneakyThrows
+    public Map<Integer, List<PdfImage>> convertImages(String dossierId, String fileId){
+
+        var imageClassificationStream = redactionStorageService.getStoredObject(RedactionStorageService.StorageIdUtils.getStorageId(dossierId, fileId, FileType.IMAGE_INFO));
+        ImageServiceResponse imageServiceResponse = objectMapper.readValue(imageClassificationStream, ImageServiceResponse.class);
+
+        Map<Integer, List<PdfImage>> images = new HashMap<>();
+        imageServiceResponse.getImageMetadata().stream().forEach(imageMetadata -> {
+            var classification = imageMetadata.getFilters().isAllPassed() ? ImageType.valueOf(imageMetadata.getClassification().getLabel().toUpperCase(Locale.ROOT)) : ImageType.OTHER;
+            images.computeIfAbsent(imageMetadata.getPosition().getPageNumber() ,x -> new ArrayList<>())
+                    .add(new PdfImage(new RedRectangle2D(imageMetadata.getPosition().getX1(), imageMetadata.getPosition().getY1(), imageMetadata.getGeometry().getWidth(), imageMetadata.getGeometry().getHeight()), classification, imageMetadata.getPosition().getPageNumber()));
+        });
+
+        return images;
+    }
+
+
+    public void findOcr(Page page){
+        page.getImages().forEach(image -> {
+            if (image.getImageType().equals(ImageType.OTHER)) {
+                page.getTextBlocks().forEach(textblock -> {
+                    if (image.getPosition()
+                            .contains(textblock.getMinX(), textblock.getMinY(), textblock.getWidth(), textblock.getHeight())) {
+                        image.setImageType(ImageType.OCR);
+                    }
+                });
+            }
+        });
+    }
+
+
+}
@@ -10,6 +10,7 @@ import java.nio.file.attribute.PosixFilePermission;
 import java.nio.file.attribute.PosixFilePermissions;
 import java.util.ArrayList;
 import java.util.List;
+import java.util.Map;
 import java.util.Set;

 import org.apache.commons.io.IOUtils;
@@ -29,7 +30,6 @@ import com.iqser.red.service.redaction.v1.server.memory.MemoryStats;
 import com.iqser.red.service.redaction.v1.server.parsing.PDFLinesTextStripper;
 import com.iqser.red.service.redaction.v1.server.parsing.model.TextPositionSequence;
 import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
-import com.iqser.red.service.redaction.v1.server.redaction.service.ImageClassificationService;
 import com.iqser.red.service.redaction.v1.server.tableextraction.model.AbstractTextContainer;
 import com.iqser.red.service.redaction.v1.server.tableextraction.model.CleanRulings;
 import com.iqser.red.service.redaction.v1.server.tableextraction.service.RulingCleaningService;
@@ -50,17 +50,10 @@ public class PdfSegmentationService {
    private final BlockificationService blockificationService;
    private final ClassificationService classificationService;
    private final SectionsBuilderService sectionsBuilderService;
-    private final ImageClassificationService imageClassificationService;
-    private final ImageMergeService imageMergeService;
+    private final ImageService imageService;


-    public Document parseDocument(InputStream documentInputStream) throws IOException {
-
-        return parseDocument(documentInputStream, false);
-    }
-
-
-    public Document parseDocument(InputStream documentInputStream, boolean ignoreImages) throws IOException {
+    public Document parseDocument(InputStream documentInputStream, Map<Integer, List<PdfImage>> pdfImages) throws IOException {

        PDDocument pdDocument = null;
        try {
@@ -122,15 +115,14 @@ public class PdfSegmentationService {
                    page.setLandscape(isLandscape || isRotated);
                    page.setPageNumber(pageNumber);

-                    List<PdfImage> mergedList = imageMergeService.mergeImages(stripper.getImages(), rotation);
-                    page.setImages(mergedList);

                    tableExtractionService.extractTables(cleanRulings, page);
                    buildPageStatistics(page);
                    increaseDocumentStatistics(page, document);

-                    if (!ignoreImages) {
-                        imageClassificationService.classifyImages(page);
+                    if (pdfImages != null && pdfImages.containsKey(pageNumber)) {
+                        page.setImages(pdfImages.get(pageNumber));
+                        imageService.findOcr(page);
                    }

                    pages.add(page);
@@ -211,7 +211,11 @@ public class SectionsBuilderService {
                if (previous != null && previous.getText().startsWith("Table ")) {
                    table.setHeadline(previous.getText());
                } else {
-                    table.setHeadline("Table in: " + lastHeadline);
+                    if(lastHeadline == null || lastHeadline.isEmpty()) {
+                        table.setHeadline("Text in table");
+                    } else {
+                        table.setHeadline("Table in: " + lastHeadline);
+                    }
                }

                if (textBlock != null && !alreadyAdded) {
@@ -1,22 +1,25 @@
 package com.iqser.red.service.redaction.v1.server.storage;

+import java.io.IOException;
+import java.io.InputStream;
+
+import org.springframework.core.io.InputStreamResource;
+import org.springframework.stereotype.Service;
+
 import com.fasterxml.jackson.databind.ObjectMapper;
 import com.iqser.red.service.persistence.service.v1.api.model.dossiertemplate.dossier.file.FileType;
 import com.iqser.red.service.redaction.v1.model.RedactionLog;
 import com.iqser.red.service.redaction.v1.model.SectionGrid;
 import com.iqser.red.service.redaction.v1.server.classification.model.Text;
 import com.iqser.red.service.redaction.v1.server.client.model.NerEntities;
+import com.iqser.red.service.redaction.v1.server.exception.NotFoundException;
 import com.iqser.red.storage.commons.exception.StorageObjectDoesNotExist;
 import com.iqser.red.storage.commons.service.StorageService;
+
 import lombok.Getter;
 import lombok.RequiredArgsConstructor;
 import lombok.SneakyThrows;
 import lombok.extern.slf4j.Slf4j;
-import org.springframework.core.io.InputStreamResource;
-import org.springframework.stereotype.Service;
-
-import java.io.IOException;
-import java.io.InputStream;

@Slf4j
@Service
@@ -26,14 +29,17 @@ public class RedactionStorageService {
    private final ObjectMapper objectMapper;
    private final StorageService storageService;

+
    @SneakyThrows
    public InputStream getStoredObject(String storageId) {
+
        return storageService.getObject(storageId).getInputStream();
    }


    @SneakyThrows
    public void storeObject(String dossierId, String fileId, FileType fileType, Object any) {
+
        storageService.storeObject(StorageIdUtils.getStorageId(dossierId, fileId, fileType), objectMapper.writeValueAsBytes(any));
    }

@@ -92,12 +98,13 @@ public class RedactionStorageService {
    }


-
    public SectionGrid getSectionGrid(String dossierId, String fileId) {

-        var sectionGrid = storageService.getObject(StorageIdUtils.getStorageId(dossierId, fileId, FileType.SECTION_GRID));
        try {
+            var sectionGrid = storageService.getObject(StorageIdUtils.getStorageId(dossierId, fileId, FileType.SECTION_GRID));
            return objectMapper.readValue(sectionGrid.getInputStream(), SectionGrid.class);
+        } catch (StorageObjectDoesNotExist e) {
+            throw new NotFoundException("Section Grid is not available.");
        } catch (IOException e) {
            throw new RuntimeException("Could not convert RedactionLog", e);
        }
@@ -116,8 +123,12 @@ public class RedactionStorageService {
    public static class StorageIdUtils {

        public static String getStorageId(String dossierId, String fileId, FileType fileType) {
+
            return dossierId + "/" + fileId + "." + fileType.name() + fileType.getExtension();
        }

+
+
    }
+
 }
@@ -24,7 +24,6 @@ import java.util.UUID;
 import java.util.stream.Collectors;

 import org.apache.commons.io.IOUtils;
-import org.apache.commons.lang3.StringUtils;
 import org.junit.After;
 import org.junit.Before;
 import org.junit.Ignore;
@@ -76,12 +75,13 @@ import com.iqser.red.service.redaction.v1.model.RedactionResult;
 import com.iqser.red.service.redaction.v1.model.StructureAnalyzeRequest;
 import com.iqser.red.service.redaction.v1.server.classification.model.SectionText;
 import com.iqser.red.service.redaction.v1.server.client.DictionaryClient;
-import com.iqser.red.service.redaction.v1.server.client.ImageClassificationClient;
 import com.iqser.red.service.redaction.v1.server.client.LegalBasisClient;
 import com.iqser.red.service.redaction.v1.server.client.RulesClient;
 import com.iqser.red.service.redaction.v1.server.controller.RedactionController;
 import com.iqser.red.service.redaction.v1.server.memory.MemoryStats;
 import com.iqser.red.service.redaction.v1.server.redaction.service.AnalyzeService;
+import com.iqser.red.service.redaction.v1.server.redaction.service.ManualRedactionSurroundingTextService;
+import com.iqser.red.service.redaction.v1.server.redaction.utils.OsUtils;
 import com.iqser.red.service.redaction.v1.server.redaction.utils.ResourceLoader;
 import com.iqser.red.service.redaction.v1.server.redaction.utils.TextNormalizationUtilities;
 import com.iqser.red.service.redaction.v1.server.storage.RedactionStorageService;
@@ -136,15 +136,15 @@ public class RedactionIntegrationTest {
    @MockBean
    private DictionaryClient dictionaryClient;

-    @MockBean
-    private ImageClassificationClient imageClassificationClient;
-
    @Autowired
    private RedactionStorageService redactionStorageService;

    @Autowired
    private StorageService storageService;

+    @Autowired
+    private ManualRedactionSurroundingTextService manualRedactionSurroundingTextService;
+
    @MockBean
    private AmazonS3 amazonS3;

@@ -591,7 +591,7 @@ public class RedactionIntegrationTest {
                .fileId(TEST_FILE_ID)
                .build());

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Annotated3.pdf")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Annotated3.pdf")) {
            fileOutputStream.write(annotateResponse.getDocument());
        }
        long rstart = System.currentTimeMillis();
@@ -607,6 +607,76 @@ public class RedactionIntegrationTest {
    }


+    @Test
+    public void redactionExpansionOverlap() throws IOException {
+
+        ClassPathResource pdfFileResource = new ClassPathResource("files/Minimal Examples/ExpansionTest.pdf");
+        AnalyzeRequest request = prepareStorage(pdfFileResource.getInputStream());
+
+        analyzeService.analyzeDocumentStructure(new StructureAnalyzeRequest(request.getDossierId(), request.getFileId()));
+        AnalyzeResult result = analyzeService.analyze(request);
+
+        var redactionLog = redactionStorageService.getRedactionLog(TEST_DOSSIER_ID, TEST_FILE_ID);
+
+        var values = redactionLog.getRedactionLogEntry()
+                .stream()
+                .map(RedactionLogEntry::getValue)
+                .collect(Collectors.toList());
+
+        assertThat(values).containsExactlyInAnyOrder("Lastname M.", "Doe", "Doe J.", "M. Mustermann", "Mustermann M.", "F. Lastname");
+    }
+
+
+    @Test
+    @SneakyThrows
+    public void testIgnoreHint() {
+
+        System.out.println("testIgnoreHint");
+
+        ClassPathResource pdfFileResource = new ClassPathResource("files/new/test-ignore-hint.pdf");
+        AnalyzeRequest request = prepareStorage(pdfFileResource.getInputStream());
+
+        analyzeService.analyzeDocumentStructure(new StructureAnalyzeRequest(request.getDossierId(), request.getFileId()));
+        analyzeService.analyze(request);
+
+        var redactionLog = redactionStorageService.getRedactionLog(TEST_DOSSIER_ID, TEST_FILE_ID);
+
+        var toRemove = IdRemoval.builder()
+                .annotationId("c630599611e6e3db314518374bcf70f7")
+                .status(AnnotationStatus.APPROVED)
+                .user("test")
+                .removeFromDictionary(false)
+                .processedDate(OffsetDateTime.now())
+                .requestDate(OffsetDateTime.now())
+                .build();
+
+        var manualRedactions = ManualRedactions.builder().idsToRemove(Set.of(toRemove)).build();
+        request.setManualRedactions(manualRedactions);
+        analyzeService.reanalyze(request);
+
+        var mergedRedactionLog = redactionController.getRedactionLog(RedactionRequest.builder()
+                .manualRedactions(manualRedactions)
+                .dossierTemplateId(TEST_DOSSIER_TEMPLATE_ID)
+                .dossierId(TEST_DOSSIER_ID)
+                .fileId(TEST_FILE_ID)
+                .build());
+
+        var cbiAddressBeforeHintRemoval = redactionLog.getRedactionLogEntry()
+                .stream()
+                .filter(re -> re.getType().equalsIgnoreCase("CBI_Address"))
+                .findAny()
+                .get();
+        assertThat(cbiAddressBeforeHintRemoval.isRedacted()).isFalse();
+
+        var cbiAddressAfterHintRemoval = mergedRedactionLog.getRedactionLogEntry()
+                .stream()
+                .filter(re -> re.getType().equalsIgnoreCase("CBI_Address"))
+                .findAny()
+                .get();
+        assertThat(cbiAddressAfterHintRemoval.isRedacted()).isTrue();
+    }
+
+
    @Test
    @Ignore
    public void noExceptionShouldBeThrownForAnyFiles() throws IOException {
@@ -664,7 +734,7 @@ public class RedactionIntegrationTest {
    public void redactionTestSeparatedRedaction() throws IOException {

        String fileName = "scanned/VV-380943_page38.pdf";
-        String outputFileName = getTemporaryDirectory() + "/AnnotatedRedactionTestSeparatedRedaction.pdf";
+        String outputFileName = OsUtils.getTemporaryDirectory() + "/AnnotatedRedactionTestSeparatedRedaction.pdf";

        long start = System.currentTimeMillis();
        ClassPathResource pdfFileResource = new ClassPathResource(fileName);
@@ -688,7 +758,7 @@ public class RedactionIntegrationTest {

        System.out.println("first analysis duration: " + (end - start));

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Test.json")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Test.json")) {
            fileOutputStream.write(objectMapper.writeValueAsBytes(redactionStorageService.getText(TEST_DOSSIER_ID, TEST_FILE_ID)));
        }

@@ -773,15 +843,60 @@ public class RedactionIntegrationTest {

        redactionLog = redactionStorageService.getRedactionLog(TEST_DOSSIER_ID, TEST_FILE_ID);

-        System.out.println("hi");
+        System.out.println("Output file:" + outputFileName);
    }

+    @Test
+    public void testChangeComputation() throws IOException {
+
+        String fileName = "files/new/test1S1T1.pdf";
+        String outputFileName = OsUtils.getTemporaryDirectory() + "/Annotated.pdf";
+
+        ClassPathResource pdfFileResource = new ClassPathResource(fileName);
+        AnalyzeRequest request = prepareStorage(pdfFileResource.getInputStream());
+
+        analyzeService.analyzeDocumentStructure(new StructureAnalyzeRequest(request.getDossierId(), request.getFileId()));
+        analyzeService.analyze(request);
+
+        dictionary.get(AUTHOR).add("report");
+        reanlysisVersions.put("report", 2L);
+        when(dictionaryClient.getVersion(TEST_DOSSIER_TEMPLATE_ID)).thenReturn(2L);
+        when(dictionaryClient.getDictionaryForType(AUTHOR +":" + TEST_DOSSIER_TEMPLATE_ID)).thenReturn(getDictionaryResponse(AUTHOR, false));
+        analyzeService.reanalyze(request);
+
+        dictionary.get(AUTHOR).add("assessment report");
+        reanlysisVersions.put("assessment report", 3L);
+        when(dictionaryClient.getVersion(TEST_DOSSIER_TEMPLATE_ID)).thenReturn(3L);
+        when(dictionaryClient.getDictionaryForType(AUTHOR +":" + TEST_DOSSIER_TEMPLATE_ID)).thenReturn(getDictionaryResponse(AUTHOR, false));
+        analyzeService.reanalyze(request);
+
+        when(dictionaryClient.getVersion(TEST_DOSSIER_TEMPLATE_ID)).thenReturn(3L);
+        when(dictionaryClient.getDictionaryForType(AUTHOR +":" + TEST_DOSSIER_TEMPLATE_ID)).thenReturn(getDictionaryResponse(AUTHOR, false));
+        analyzeService.reanalyze(request);
+
+        var  redactionLog = redactionStorageService.getRedactionLog(TEST_DOSSIER_ID, TEST_FILE_ID);
+
+        var changes = redactionLog.getRedactionLogEntry().stream().filter(entry ->
+                entry.getValue() != null && entry.getValue().equals("report"))
+                .findFirst().get().getChanges();
+
+        assertThat(changes.size()).isEqualTo(2);
+
+        AnnotateResponse annotateResponse = redactionController.annotate(AnnotateRequest.builder()
+                .dossierId(TEST_DOSSIER_ID)
+                .fileId(TEST_FILE_ID)
+                .build());
+
+        try (FileOutputStream fileOutputStream = new FileOutputStream(outputFileName)) {
+            fileOutputStream.write(annotateResponse.getDocument());
+        }
+    }

    @Test
    public void redactionTest() throws IOException {

-        String fileName = "files/new/VV-919901.pdf";
-        String outputFileName = getTemporaryDirectory() + "/Annotated.pdf";
+        String fileName = "files/new/Single Study - Oral (Gavage) Mouse.pdf";
+        String outputFileName = OsUtils.getTemporaryDirectory() + "/Annotated.pdf";

        long start = System.currentTimeMillis();
        ClassPathResource pdfFileResource = new ClassPathResource(fileName);
@@ -805,7 +920,7 @@ public class RedactionIntegrationTest {

        System.out.println("first analysis duration: " + (end - start));

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Test.json")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Test.json")) {
            fileOutputStream.write(objectMapper.writeValueAsBytes(redactionStorageService.getText(TEST_DOSSIER_ID, TEST_FILE_ID)));
        }

@@ -890,7 +1005,7 @@ public class RedactionIntegrationTest {

        redactionLog = redactionStorageService.getRedactionLog(TEST_DOSSIER_ID, TEST_FILE_ID);

-        System.out.println("hi");
+        System.out.println("Output file:" + outputFileName);
    }


@@ -909,7 +1024,7 @@ public class RedactionIntegrationTest {
                .fileId(TEST_FILE_ID)
                .build());

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Annotated.pdf")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Annotated.pdf")) {
            fileOutputStream.write(annotateResponse.getDocument());
        }
        long end = System.currentTimeMillis();
@@ -1003,7 +1118,7 @@ public class RedactionIntegrationTest {
                .fileId(TEST_FILE_ID)
                .build());

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Annotated.pdf")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Annotated.pdf")) {
            fileOutputStream.write(annotateResponse.getDocument());
        }
        long end = System.currentTimeMillis();
@@ -1029,7 +1144,7 @@ public class RedactionIntegrationTest {

        RedactionResult result = redactionController.classify(redactionRequest);

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Classified.pdf")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Classified.pdf")) {
            fileOutputStream.write(result.getDocument());
        }
    }
@@ -1051,7 +1166,7 @@ public class RedactionIntegrationTest {

        RedactionResult result = redactionController.sections(redactionRequest);

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Sections.pdf")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Sections.pdf")) {
            fileOutputStream.write(result.getDocument());
        }
    }
@@ -1073,7 +1188,7 @@ public class RedactionIntegrationTest {

        RedactionResult result = redactionController.htmlTables(redactionRequest);

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Tables.html")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Tables.html")) {
            fileOutputStream.write(result.getDocument());
        }
    }
@@ -1095,7 +1210,7 @@ public class RedactionIntegrationTest {

        RedactionResult result = redactionController.htmlTables(redactionRequest);

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Tables.html")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Tables.html")) {
            fileOutputStream.write(result.getDocument());
        }
    }
@@ -1165,7 +1280,7 @@ public class RedactionIntegrationTest {
                .fileId(TEST_FILE_ID)
                .build());

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Annotated.pdf")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Annotated.pdf")) {
            fileOutputStream.write(annotateResponse.getDocument());
        }
        long end = System.currentTimeMillis();
@@ -1298,12 +1413,13 @@ public class RedactionIntegrationTest {
                .manualRedactions(manualRedactions)
                .build());

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Annotated.pdf")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Annotated.pdf")) {
            fileOutputStream.write(annotateResponse.getDocument());
        }

    }

+
    @Test
    public void testExpandByRegEx() throws IOException {

@@ -1319,7 +1435,7 @@ public class RedactionIntegrationTest {
                .fileId(TEST_FILE_ID)
                .build());

-        try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Annotated.pdf")) {
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Annotated.pdf")) {
            fileOutputStream.write(annotateResponse.getDocument());
        }
        long end = System.currentTimeMillis();
@@ -1329,6 +1445,86 @@ public class RedactionIntegrationTest {
    }


+    @Test
+    public void testManualSurroundingText() throws IOException {
+
+        ClassPathResource pdfFileResource = new ClassPathResource("files/new/S4.pdf");
+
+        ManualRedactions manualRedactions = new ManualRedactions();
+
+        ManualRedactionEntry manualRedactionEntry = new ManualRedactionEntry();
+        manualRedactionEntry.setAnnotationId(UUID.randomUUID().toString());
+        manualRedactionEntry.setFileId("fileId");
+        manualRedactionEntry.setStatus(AnnotationStatus.APPROVED);
+        manualRedactionEntry.setType("CBI_author");
+        manualRedactionEntry.setValue("rabbits");
+        manualRedactionEntry.setReason("Manual Redaction");
+        manualRedactionEntry.setPositions(List.of(Rectangle.builder()
+                .topLeftX(70.944f)
+                .topLeftY(670.1595f)
+                .width(30.07296f)
+                .height(10.048125f)
+                .page(1)
+                .build()));
+
+        ManualRedactionEntry manualRedactionEntry2 = new ManualRedactionEntry();
+        manualRedactionEntry2.setAnnotationId(UUID.randomUUID().toString());
+        manualRedactionEntry2.setFileId("fileId");
+        manualRedactionEntry2.setStatus(AnnotationStatus.APPROVED);
+        manualRedactionEntry2.setType("CBI_author");
+        manualRedactionEntry2.setValue("rabbits");
+        manualRedactionEntry2.setReason("Manual Redaction");
+        manualRedactionEntry2.setPositions(List.of(Rectangle.builder()
+                .topLeftX(470.5204f)
+                .topLeftY(746.1195f)
+                .width(29.96256f)
+                .height(10.048125f)
+                .page(1)
+                .build()));
+
+        ManualRedactionEntry manualRedactionEntry3 = new ManualRedactionEntry();
+        manualRedactionEntry3.setAnnotationId(UUID.randomUUID().toString());
+        manualRedactionEntry3.setFileId("fileId");
+        manualRedactionEntry3.setStatus(AnnotationStatus.APPROVED);
+        manualRedactionEntry3.setType("CBI_author");
+        manualRedactionEntry3.setValue("AOEL");
+        manualRedactionEntry3.setReason("Manual Redaction");
+        manualRedactionEntry3.setPositions(List.of(Rectangle.builder()
+                .topLeftX(355.53775f)
+                .topLeftY(266.1895f)
+                .width(29.32224f)
+                .height(10.048125f)
+                .page(1)
+                .build()));
+
+        manualRedactions.getEntriesToAdd().add(manualRedactionEntry);
+        manualRedactions.getEntriesToAdd().add(manualRedactionEntry2);
+        manualRedactions.getEntriesToAdd().add(manualRedactionEntry3);
+
+        AnalyzeRequest request = prepareStorage(pdfFileResource.getInputStream());
+        request.setManualRedactions(manualRedactions);
+        analyzeService.analyzeDocumentStructure(new StructureAnalyzeRequest(request.getDossierId(), request.getFileId()));
+        AnalyzeResult result = analyzeService.analyze(request);
+
+        AnnotateResponse annotateResponse = redactionController.annotate(AnnotateRequest.builder()
+                .dossierId(TEST_DOSSIER_ID)
+                .dossierTemplateId(TEST_DOSSIER_TEMPLATE_ID)
+                .fileId(TEST_FILE_ID)
+                .manualRedactions(manualRedactions)
+                .build());
+
+        try (FileOutputStream fileOutputStream = new FileOutputStream(OsUtils.getTemporaryDirectory() + "/Annotated.pdf")) {
+            fileOutputStream.write(annotateResponse.getDocument());
+        }
+
+        var surroundingTextResult = manualRedactionSurroundingTextService.addSurroundingText(TEST_DOSSIER_ID, TEST_FILE_ID, manualRedactions);
+        surroundingTextResult.getEntriesToAdd().forEach(addEntry -> {
+            assertThat(addEntry.getTextAfter()).isNotEmpty();
+        });
+
+    }
+
+
    private static String loadFromClassPath(String path) {

        URL resource = ResourceLoader.class.getClassLoader().getResource(path);
@@ -1365,14 +1561,4 @@ public class RedactionIntegrationTest {

    }

-
-    private static String getTemporaryDirectory() {
-
-        String tmpdir = System.getProperty("java.io.tmpdir");
-        if (StringUtils.isNotBlank(tmpdir)) {
-            return tmpdir;
-        }
-        return "/tmp";
-    }
-
-}
+}
@@ -151,8 +151,7 @@ public class EntitySearchUtilsTest {
        Set<Entity> result = EntitySearchUtils.findNonOverlappingMatchEntities(existingEntities, foundEntities);

        // Assert
-        assertThat(result.size()).isEqualTo(1);
-        assertThat(result).contains(foundEntities2);
+        assertThat(result.size()).isEqualTo(0);

    }

@@ -182,8 +181,7 @@ public class EntitySearchUtilsTest {
        Set<Entity> result = EntitySearchUtils.findNonOverlappingMatchEntities(existingEntities, foundEntities);

        // Assert
-        assertThat(result.size()).isEqualTo(1);
-        assertThat(result).contains(foundEntities2);
+        assertThat(result.size()).isEqualTo(0);

    }

@@ -0,0 +1,22 @@
+package com.iqser.red.service.redaction.v1.server.redaction.utils;
+
+import org.apache.commons.lang3.StringUtils;
+
+public class OsUtils {
+
+    public static boolean isWindows() {
+
+        return StringUtils.containsIgnoreCase(System.getProperty("os.name"), "Windows");
+    }
+
+
+    public static String getTemporaryDirectory() {
+
+        String tmpdir = System.getProperty("java.io.tmpdir");
+        if (isWindows() && StringUtils.isNotBlank(tmpdir)) {
+            return tmpdir;
+        }
+        return "/tmp";
+    }
+
+}
@@ -1,12 +1,16 @@
 package com.iqser.red.service.redaction.v1.server.segmentation;

 import com.amazonaws.services.s3.AmazonS3;
+import com.fasterxml.jackson.databind.ObjectMapper;
 import com.iqser.red.service.redaction.v1.server.Application;
 import com.iqser.red.service.redaction.v1.server.classification.model.Document;
 import com.iqser.red.service.redaction.v1.server.classification.model.Page;
 import com.iqser.red.service.redaction.v1.server.classification.service.BlockificationService;
 import com.iqser.red.service.redaction.v1.server.client.LegalBasisClient;
+import com.iqser.red.service.redaction.v1.server.redaction.model.ImageType;
 import com.iqser.red.service.redaction.v1.server.redaction.model.PdfImage;
+import com.iqser.red.service.redaction.v1.server.redaction.model.RedRectangle2D;
+import com.iqser.red.service.redaction.v1.server.redaction.model.image.ImageServiceResponse;
 import com.iqser.red.service.redaction.v1.server.tableextraction.model.Cell;
 import com.iqser.red.service.redaction.v1.server.tableextraction.model.Table;
 import com.iqser.red.service.redaction.v1.server.tableextraction.service.RulingCleaningService;
@@ -31,12 +35,18 @@ import javax.imageio.ImageIO;
 import java.io.ByteArrayOutputStream;
 import java.io.FileOutputStream;
 import java.io.IOException;
+import java.util.ArrayList;
 import java.util.Collections;
+import java.util.HashMap;
 import java.util.List;
+import java.util.Locale;
+import java.util.Map;
 import java.util.stream.Collectors;

+import static com.iqser.red.service.redaction.v1.server.redaction.utils.OsUtils.getTemporaryDirectory;
 import static org.assertj.core.api.Assertions.assertThat;

+import lombok.SneakyThrows;

@RunWith(SpringRunner.class)
@SpringBootTest(classes = Application.class, webEnvironment = SpringBootTest.WebEnvironment.RANDOM_PORT)
@@ -67,42 +77,29 @@ public class PdfSegmentationServiceTest {
    @MockBean
    private LegalBasisClient legalBasisClient;

+    @Autowired
+    private ObjectMapper objectMapper;
+
    @Configuration
    @EnableAutoConfiguration(exclude = { RabbitAutoConfiguration.class})
    public static class TestConfiguration {

    }

-    @Test
-    public void testMergeImages() throws IOException {
-
-        ClassPathResource pdfFileResource = new ClassPathResource("files/Minimal Examples/270Rotated.pdf");
-
-        Document document = pdfSegmentationService.parseDocument(pdfFileResource.getInputStream());
-        assertThat(document.getPages().get(0).getImages().size()).isEqualTo(1);
-        assertThat(document.getPages().get(1).getImages().size()).isEqualTo(0);
-
-    }

    @Test
-    @Ignore
-    public void testExtractImages() throws IOException {
+    @SneakyThrows
+    public void testMapping(){
+        ClassPathResource responseJson = new ClassPathResource("files/image_response.json");
+        ImageServiceResponse imageServiceResponse = objectMapper.readValue(responseJson.getInputStream(), ImageServiceResponse.class);

-        ClassPathResource pdfFileResource = new ClassPathResource("files/new/Single Study - Oral (Gavage) Mouse.pdf");
+        Map<Integer, List<PdfImage>> images = new HashMap<>();
+        imageServiceResponse.getImageMetadata().stream().forEach(imageMetadata -> {
+            images.computeIfAbsent(imageMetadata.getPosition().getPageNumber() ,x -> new ArrayList<>())
+                    .add(new PdfImage(new RedRectangle2D(imageMetadata.getPosition().getX1(), imageMetadata.getPosition().getY1(), imageMetadata.getGeometry().getWidth(), imageMetadata.getGeometry().getHeight()), ImageType.valueOf(imageMetadata.getClassification().getLabel().toUpperCase(Locale.ROOT)), imageMetadata.getPosition().getPageNumber()));
+        });

-        Document document = pdfSegmentationService.parseDocument(pdfFileResource.getInputStream());
-        int i = 0;
-        for (Page page : document.getPages()) {
-            for (PdfImage image : page.getImages()) {
-                try (ByteArrayOutputStream baos = new ByteArrayOutputStream()) {
-                    ImageIO.write(image.getImage(), "png", baos);
-                    try (FileOutputStream fileOutputStream = new FileOutputStream(getTemporaryDirectory() + "/Image " + i + ".png")) {
-                        fileOutputStream.write(baos.toByteArray());
-                    }
-                }
-                i++;
-            }
-        }
+        System.out.println("object");
    }


@@ -111,7 +108,7 @@ public class PdfSegmentationServiceTest {

        ClassPathResource pdfFileResource = new ClassPathResource("files/Minimal Examples/Spanning Cells.pdf");

-        Document document = pdfSegmentationService.parseDocument(pdfFileResource.getInputStream());
+        Document document = pdfSegmentationService.parseDocument(pdfFileResource.getInputStream(), null);
        assertThat(document.getParagraphs()
                .stream()
                .flatMap(paragraph -> paragraph.getTables().stream())
@@ -132,7 +129,7 @@ public class PdfSegmentationServiceTest {

        ClassPathResource pdfFileResource = new ClassPathResource("files/Minimal Examples/Merge Table.pdf");

-        Document document = pdfSegmentationService.parseDocument(pdfFileResource.getInputStream());
+        Document document = pdfSegmentationService.parseDocument(pdfFileResource.getInputStream(), null);
        assertThat(document.getParagraphs()
                .stream()
                .flatMap(paragraph -> paragraph.getTables().stream())
@@ -170,7 +167,7 @@ public class PdfSegmentationServiceTest {

        ClassPathResource pdfFileResource = new ClassPathResource("files/Minimal Examples/Merge Multi Page Table.pdf");

-        Document document = pdfSegmentationService.parseDocument(pdfFileResource.getInputStream());
+        Document document = pdfSegmentationService.parseDocument(pdfFileResource.getInputStream(), null);
        assertThat(document.getParagraphs()
                .stream()
                .flatMap(paragraph -> paragraph.getTables().stream())
@@ -208,7 +205,7 @@ public class PdfSegmentationServiceTest {

        ClassPathResource pdfFileResource = new ClassPathResource("files/Minimal Examples/Rotated Table Headers.pdf");

-        Document document = pdfSegmentationService.parseDocument(pdfFileResource.getInputStream());
+        Document document = pdfSegmentationService.parseDocument(pdfFileResource.getInputStream(), null);
        assertThat(document.getParagraphs()
                .stream()
                .flatMap(paragraph -> paragraph.getTables().stream())
@@ -240,12 +237,5 @@ public class PdfSegmentationServiceTest {
                        .equals(firstTableHeaderCells))).isTrue();
    }

-    private static String getTemporaryDirectory() {
-        String tmpdir = System.getProperty("java.io.tmpdir");
-        if (StringUtils.isNotBlank(tmpdir)) {
-            return tmpdir;
-        }
-        return "/tmp";
-    }

 }
@@ -1660,3 +1660,4 @@ Zyma SA
 Zyma SA, Nyon, Switzerland
 Mambo-Tox Ltd. Biomedical Sciences Building Bassett Crescent East Southampton SO16 7PX UK
 Syngenta Environmental Sciences Jealott’s Hill International Research Centre Bracknell, Berkshire RG42 6EY UK
+Test Ignored Hint CBI_ADDRESS
@@ -8615,4 +8615,9 @@ Zoriki Hosomi R.
 Zoriki Hosomi Rosana
 Zuberer D
 Zubrod J
-Zwicker R.E.
+Zwicker R.E.
+Doe
+M. Mustermann
+F. Lastname
+Mustermann
+Lastname
@@ -85,3 +85,4 @@ Toxicology and Applied Pharmacology
 Toxicol Sci
 Toxicol Sci.
 Toxicol Sci. 1
+Test Ignored Hint Published Information
@@ -19,8 +19,8 @@ rule "0: Expand CBI Authors with firstname initials"
    when
        Section(matchesType("CBI_author") || matchesType("recommendation_CBI_author"))
    then
-        section.expandByRegEx("CBI_author", "(,? [A-Z]\\.?( ?[A-Z]\\.?)?( ?[A-Z]\\.?)?\\b\\.?)", false, 1, "[\\s]+");
-        section.expandByRegEx("recommendation_CBI_author", "(,? [A-Z]\\.?( ?[A-Z]\\.?)?( ?[A-Z]\\.?)?\\b\\.?)", false, 1, "[\\s]+");
+        section.expandByRegEx("CBI_author", "(,? [A-Z]\\.?( ?[A-Z]\\.?)?( ?[A-Z]\\.?)?\\b\\.?)", false, 1, "[^\\s]+");
+        section.expandByRegEx("recommendation_CBI_author", "(,? [A-Z]\\.?( ?[A-Z]\\.?)?( ?[A-Z]\\.?)?\\b\\.?)", false, 1, "[^\\s]+");
    end


@@ -0,0 +1,686 @@
+{
+  "dossierId": "f889853e-4bf8-49a9-aae5-c38605c6ef40",
+  "fileId": "22ef63e29bb2a27db8497272336f6b32",
+  "imageMetadata": [
+    {
+      "classification": {
+        "probabilities": {
+          "logo": 1.0,
+          "signature": 0.0,
+          "other": 0.0,
+          "formula": 0.0
+        },
+        "label": "logo"
+      },
+      "position": {
+        "x1": 89.88,
+        "x2": 274.20000000000005,
+        "y1": 716.24,
+        "y2": 770.0,
+        "pageNumber": 1
+      },
+      "geometry": {
+        "width": 184.32000000000005,
+        "height": 53.75999999999999
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.14298074612038092,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 3.42857142857143,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "signature": 0.999968,
+          "logo": 1.6e-05,
+          "other": 1.6e-05,
+          "formula": 0.0
+        },
+        "label": "signature"
+      },
+      "position": {
+        "x1": -0.10000600000000001,
+        "x2": 595.099994,
+        "y1": -0.07998660000000002,
+        "y2": 842.0800134,
+        "pageNumber": 3
+      },
+      "geometry": {
+        "width": 595.2,
+        "height": 842.16
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 1.0000782051152328,
+            "tooLarge": true,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.706754060986036,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": false
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "signature": 0.999872,
+          "other": 7.9e-05,
+          "logo": 4.8e-05,
+          "formula": 0.0
+        },
+        "label": "signature"
+      },
+      "position": {
+        "x1": -0.10000600000000001,
+        "x2": 595.099994,
+        "y1": -0.07998660000000002,
+        "y2": 842.0800134,
+        "pageNumber": 7
+      },
+      "geometry": {
+        "width": 595.2,
+        "height": 842.16
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 1.0000782051152328,
+            "tooLarge": true,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.706754060986036,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": false
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "signature": 0.996366,
+          "other": 0.00,
+          "logo": 2.3e-05,
+          "formula": 4e-06
+        },
+        "label": "signature"
+      },
+      "position": {
+        "x1": -0.10000600000000001,
+        "x2": 595.099994,
+        "y1": -0.07998660000000002,
+        "y2": 842.0800134,
+        "pageNumber": 8
+      },
+      "geometry": {
+        "width": 595.2,
+        "height": 842.16
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 1.0002630764355351,
+            "tooLarge": true,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.706754060986036,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": false
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "signature": 0.999772,
+          "logo": 0.000131,
+          "other": 9.7e-05,
+          "formula": 0.0
+        },
+        "label": "signature"
+      },
+      "position": {
+        "x1": 82.59443842482001,
+        "x2": 512.6365568843402,
+        "y1": 116.943736387567,
+        "y2": 725.0718450317352,
+        "pageNumber": 73
+      },
+      "geometry": {
+        "width": 430.04211845952017,
+        "height": 608.1281086441682
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.72236755521117,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.7071571143427432,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "signature": 1.0,
+          "logo": 0.0,
+          "other": 0.0,
+          "formula": 0.0
+        },
+        "label": "signature"
+      },
+      "position": {
+        "x1": 328.20483600000006,
+        "x2": 393.94460940000005,
+        "y1": 175.1643178,
+        "y2": 203.92865619999998,
+        "pageNumber": 81
+      },
+      "geometry": {
+        "width": 65.73977339999999,
+        "height": 28.764338399999986
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.06142518774572455,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 2.2854609929078022,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "signature": 1.0,
+          "logo": 0.0,
+          "other": 0.0,
+          "formula": 0.0
+        },
+        "label": "signature"
+      },
+      "position": {
+        "x1": 136.5955818,
+        "x2": 224.72461859999999,
+        "y1": 175.1133172,
+        "y2": 203.97965680000001,
+        "pageNumber": 81
+      },
+      "geometry": {
+        "width": 88.1290368,
+        "height": 28.866339600000003
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.07124601312700823,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 3.053003533568904,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "logo": 1.0,
+          "signature": 0.0,
+          "other": 0.0,
+          "formula": 0.0
+        },
+        "label": "logo"
+      },
+      "position": {
+        "x1": 194.99126880000003,
+        "x2": 399.80967840000005,
+        "y1": 554.6597824,
+        "y2": 686.2413304,
+        "pageNumber": 81
+      },
+      "geometry": {
+        "width": 204.81840960000002,
+        "height": 131.581548
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.23189275858788796,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 1.5565891472868219,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "other": 1.0,
+          "signature": 0.0,
+          "formula": 0.0,
+          "logo": 0.0
+        },
+        "label": "other"
+      },
+      "position": {
+        "x1": 133.9945512,
+        "x2": 242.52382799999998,
+        "y1": 411.24609519999996,
+        "y2": 523.2434128,
+        "pageNumber": 90
+      },
+      "geometry": {
+        "width": 108.52927679999999,
+        "height": 111.99731760000003
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.15573364968831904,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.9690346083788703,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "other": 1.0,
+          "signature": 0.0,
+          "formula": 0.0,
+          "logo": 0.0
+        },
+        "label": "other"
+      },
+      "position": {
+        "x1": 133.5865464,
+        "x2": 242.3198256,
+        "y1": 274.972492,
+        "y2": 387.7858192,
+        "pageNumber": 90
+      },
+      "geometry": {
+        "width": 108.7332792,
+        "height": 112.8133272
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.15644678522591335,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.9638336347197106,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "other": 1.0,
+          "signature": 0.0,
+          "formula": 0.0,
+          "logo": 0.0
+        },
+        "label": "other"
+      },
+      "position": {
+        "x1": 246.19587120000003,
+        "x2": 356.5611696,
+        "y1": 400.84197279999995,
+        "y2": 519.3673672,
+        "pageNumber": 90
+      },
+      "geometry": {
+        "width": 110.3652984,
+        "height": 118.52539440000004
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.1615575178049721,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.9311531841652321,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "other": 1.0,
+          "signature": 0.0,
+          "formula": 0.0,
+          "logo": 0.0
+        },
+        "label": "other"
+      },
+      "position": {
+        "x1": 244.9718568,
+        "x2": 358.3971912,
+        "y1": 274.972492,
+        "y2": 387.7858192,
+        "pageNumber": 90
+      },
+      "geometry": {
+        "width": 113.4253344,
+        "height": 112.8133272
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.15978662903260646,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 1.0054249547920433,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "other": 1.0,
+          "signature": 0.0,
+          "formula": 0.0,
+          "logo": 0.0
+        },
+        "label": "other"
+      },
+      "position": {
+        "x1": 254.9679744,
+        "x2": 371.6573472,
+        "y1": 439.6024288,
+        "y2": 564.0438928,
+        "pageNumber": 91
+      },
+      "geometry": {
+        "width": 116.6893728,
+        "height": 124.441464
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.17021718544102565,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.9377049180327869,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "other": 1.0,
+          "signature": 0.0,
+          "formula": 0.0,
+          "logo": 0.0
+        },
+        "label": "other"
+      },
+      "position": {
+        "x1": 133.9945512,
+        "x2": 249.663912,
+        "y1": 443.07046959999997,
+        "y2": 687.2613424,
+        "pageNumber": 91
+      },
+      "geometry": {
+        "width": 115.66936080000002,
+        "height": 244.19087280000002
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.23739910530627284,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.4736842105263158,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "other": 1.0,
+          "signature": 0.0,
+          "formula": 0.0,
+          "logo": 0.0
+        },
+        "label": "other"
+      },
+      "position": {
+        "x1": 105.84222,
+        "x2": 374.870385,
+        "y1": 526.40545,
+        "y2": 687.05734,
+        "pageNumber": 92
+      },
+      "geometry": {
+        "width": 269.028165,
+        "height": 160.65188999999998
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.2936614851112628,
+            "tooLarge": false,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 1.6746031746031749,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": true
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "logo": 0.788068,
+          "other": 0.152259,
+          "formula": 0.036883,
+          "signature": 0.02279
+        },
+        "label": "logo"
+      },
+      "position": {
+        "x1": 44.64999049990001,
+        "x2": 550.5759424999001,
+        "y1": 63.286004150029996,
+        "y2": 778.72242095003,
+        "pageNumber": 94
+      },
+      "geometry": {
+        "width": 505.92595200000005,
+        "height": 715.4364168000001
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.8498341845521462,
+            "tooLarge": true,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.7071571143427431,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": false
+      }
+    },
+    {
+      "classification": {
+        "probabilities": {
+          "signature": 0.998335,
+          "logo": 0.000955,
+          "other": 0.000703,
+          "formula": 7e-06
+        },
+        "label": "signature"
+      },
+      "position": {
+        "x1": 58.954005540029996,
+        "x2": 536.45979618003,
+        "y1": 83.94401504006001,
+        "y2": 758.05854296006,
+        "pageNumber": 95
+      },
+      "geometry": {
+        "width": 477.50579064000004,
+        "height": 674.11452792
+      },
+      "filters": {
+        "geometry": {
+          "imageSize": {
+            "quotient": 0.8014221863697041,
+            "tooLarge": true,
+            "tooSmall": false
+          },
+          "imageFormat": {
+            "quotient": 0.7083452007974936,
+            "tooTall": false,
+            "tooWide": false
+          }
+        },
+        "probability": {
+          "unconfident": false
+        },
+        "allPassed": false
+      }
+    }
+  ]
+}
Author	SHA1	Message	Date
Dominique Eiflaender	9c9583067f	Pull request #316 : RED-2440: Integrated image-service-v2 Merge in RED/redaction-service from RED-2440 to master * commit 'dee1aa1f01085e585bb3126b7883bd3d01090f17': RED-2440: Integrated image-service-v2	2021-12-22 10:40:04 +01:00
deiflaender	dee1aa1f01	RED-2440: Integrated image-service-v2	2021-12-22 10:35:13 +01:00
Ali Oezyetimoglu	db16f8c1da	Pull request #314 : RED-3127: Upgrade log4j to 2.17.0 Merge in RED/redaction-service from RED-3127-rs1 to master * commit '6911c5cdd5a0e559f4abfbd083b29aca9fbae9d9': RED-3127: Upgrade log4j to 2.17.0	2021-12-21 12:17:40 +01:00
aoezyetimoglu	6911c5cdd5	RED-3127: Upgrade log4j to 2.17.0	2021-12-21 12:00:24 +01:00
Timo Bejan	37c3814ebd	Pull request #313 : condition fix Merge in RED/redaction-service from redaction-log-merge-service-test to master * commit 'e8ae8728f3abd82cacd641144406f9047ca9c262': condition fix	2021-12-21 11:07:37 +01:00
Timo Bejan	e8ae8728f3	condition fix	2021-12-21 11:18:37 +02:00
Timo Bejan	b0415adaf4	Pull request #312 : Redaction log merge service test Merge in RED/redaction-service from redaction-log-merge-service-test to master * commit '9e64773d0db262481f8552678a6e6b86d81d93c4': pmd skip remove from dict entries from merge	2021-12-21 10:17:54 +01:00
Timo Bejan	9e64773d0d	pmd	2021-12-21 11:08:15 +02:00
Timo Bejan	3145195d63	skip remove from dict entries from merge	2021-12-21 11:05:47 +02:00
Dominique Eiflaender	88b1274cd3	Pull request #311 : RED-3117: Enbled to ignore invalid entries at addToDictionary for automated pushing entries in analysis Merge in RED/redaction-service from RED-3117 to master * commit '0296745efc0e4786386811d0037e2cae6ae511c4': RED-3117: Enbled to ignore invalid entries at addToDictionary for automated pushing entries in analysis	2021-12-20 11:58:34 +01:00
deiflaender	0296745efc	RED-3117: Enbled to ignore invalid entries at addToDictionary for automated pushing entries in analysis	2021-12-20 11:53:36 +01:00
Dominique Eiflaender	60a0adb843	Pull request #310 : RED-3114: Fixed unreliable redactions on rotated pages Merge in RED/redaction-service from RED-3114 to master * commit 'f1071de3a9ea9a3b34798f21ee9589262ef4f32b': RED-3114: Fixed unreliable redactions on rotated pages	2021-12-17 16:00:18 +01:00
deiflaender	f1071de3a9	RED-3114: Fixed unreliable redactions on rotated pages	2021-12-17 15:55:09 +01:00
Ali Oezyetimoglu	681d17ad95	Pull request #309 : RED-3116: Fix missing relativePath in mvn projects to leads to non working sync on lastest intellij Merge in RED/redaction-service from RED-3116-rs1 to master * commit '8deb24a5342c1c7a34fa6235e3d68f8161d884e4': RED-3116: Fix missing relativePath in mvn projects to leads to non working sync on lastest intellij	2021-12-17 13:53:14 +01:00
aoezyetimoglu	8deb24a534	RED-3116: Fix missing relativePath in mvn projects to leads to non working sync on lastest intellij	2021-12-17 13:10:43 +01:00
Ali Oezyetimoglu	98443f50ec	Pull request #308 : RED-3107 rs1 Merge in RED/redaction-service from RED-3107-rs1 to master * commit '7d75961df45de93cfb2a708b640703dbd5693f40': RED-3107: Missing headline for tables appears as inconsistent text in report RED-3106: Upgrade log4j to 2.16.0	2021-12-17 12:46:53 +01:00
aoezyetimoglu	7d75961df4	RED-3107: Missing headline for tables appears as inconsistent text in report	2021-12-17 12:00:14 +01:00
aoezyetimoglu	3c1f4b4853	Merge branch 'master' of ssh://git.iqser.com:2222/red/redaction-service	2021-12-17 11:52:52 +01:00
Dominique Eiflaender	c9e74f4ff9	Pull request #307 : RED-3105: Fixed change computation Merge in RED/redaction-service from RED-3105 to master * commit '19dcf0a67434442b2dd578523bbc5efa49c644e8': RED-3105: Fixed change computation	2021-12-16 14:22:55 +01:00
deiflaender	19dcf0a674	RED-3105: Fixed change computation	2021-12-16 14:17:04 +01:00
Philipp Schramm	c46c41b790	Pull request #306 : RED-3067 Bugfix to show redactions on rotated pages correctly Merge in RED/redaction-service from RED-3067 to master * commit 'f49782438461ec34e6e8c0ccf88b3017755e593d': RED-3067 Bugfix to show redactions on rotated pages correctly and fixed getting tmp directory RED-3067 Bugfix to show redactions on rotated pages correctly	2021-12-16 14:14:40 +01:00
Philipp Schramm	f497824384	RED-3067 Bugfix to show redactions on rotated pages correctly and fixed getting tmp directory	2021-12-16 14:08:05 +01:00
Philipp Schramm	f7fbe8f0f6	RED-3067 Bugfix to show redactions on rotated pages correctly	2021-12-16 12:43:17 +01:00
Ali Oezyetimoglu	e8dce509f3	Pull request #303 : RED-3106: Upgrade log4j to 2.16.0 Merge in RED/redaction-service from RED-3106-rs1 to master * commit '9346cc701e1601b6c9d44e41de9fc193fbc81b1e': RED-3106: Upgrade log4j to 2.16.0	2021-12-16 09:57:54 +01:00
aoezyetimoglu	b125de2b55	RED-3106: Upgrade log4j to 2.16.0	2021-12-16 09:52:21 +01:00
aoezyetimoglu	9346cc701e	RED-3106: Upgrade log4j to 2.16.0	2021-12-16 09:44:04 +01:00
Dominique Eiflaender	1987e7b430	Pull request #302 : RED-3025: Ported code to ignore hints Merge in RED/redaction-service from RED-3025 to master * commit '010a3ee6ab7f06297c2c37691bcac2f7019fb3a0': RED-3025: Ported code to ignore hints	2021-12-15 13:26:48 +01:00
deiflaender	010a3ee6ab	RED-3025: Ported code to ignore hints	2021-12-15 13:22:28 +01:00
Philipp Schramm	358f8814cd	Pull request #300 : RED-3032 Bugfix: Throw NotFoundException in case section grid was not found Merge in RED/redaction-service from bugfix/RED-3032 to master * commit '9fb4024fb7dfe7ae3f92f5c17decf15fbac58451': RED-3032 Bugfix: Throw NotFoundException in case section grid was not found	2021-12-15 09:05:10 +01:00
Philipp Schramm	9fb4024fb7	RED-3032 Bugfix: Throw NotFoundException in case section grid was not found	2021-12-14 17:02:24 +01:00
Ali Oezyetimoglu	a155f18adc	Pull request #299 : RED-3061 rs3 Merge in RED/redaction-service from RED-3061-rs3 to master * commit '1ccf0ab493fcb367a071641e2fea050cf6ad26cf': RED-3061: Logic for value matcher for initials expansion rule is inverted RED-3061: Logic for value matcher for initials expansion rule is inverted	2021-12-14 12:35:45 +01:00
aoezyetimoglu	1ccf0ab493	RED-3061: Logic for value matcher for initials expansion rule is inverted	2021-12-14 12:25:21 +01:00
aoezyetimoglu	1b9251b0a2	RED-3061: Logic for value matcher for initials expansion rule is inverted	2021-12-14 11:10:59 +01:00
Dominique Eiflaender	6bfa54583f	Pull request #297 : RED-2439: Fixed missing MessageType in AnalyseResult Merge in RED/redaction-service from RED-2439-4 to master * commit '83a757976adf26cfeae8fb682593ce1d5f606dfc': RED-2439: Fixed missing MessageType in AnalyseResult	2021-12-13 13:55:52 +01:00
deiflaender	83a757976a	RED-2439: Fixed missing MessageType in AnalyseResult	2021-12-13 13:51:33 +01:00
Dominique Eiflaender	bd5c997572	Pull request #296 : RED-2439: Enabled to find surrounding text for manual redactions via queue Merge in RED/redaction-service from RED-2439-3 to master * commit 'e4cf3bfa602014cd8a2f2162199557880600e93f': RED-2439: Enabled to find surrounding text for manual redactions via queue	2021-12-13 12:32:20 +01:00
deiflaender	e4cf3bfa60	RED-2439: Enabled to find surrounding text for manual redactions via queue	2021-12-13 12:24:26 +01:00
Ali Oezyetimoglu	3f153c1df5	Pull request #294 : RED-3075: Upgraded log4j to 2.15.0 Merge in RED/redaction-service from RED-3075-rs2 to master * commit '142c81d27d9e5e67532b3d8795c2d4c559342d37': RED-3075: Upgraded log4j to 2.15.0	2021-12-13 11:03:56 +01:00
aoezyetimoglu	142c81d27d	RED-3075: Upgraded log4j to 2.15.0	2021-12-13 10:59:16 +01:00
Dominique Eiflaender	a2c57b94a2	Pull request #292 : RED-3059: Fixed inital expansion overlaps Merge in RED/redaction-service from RED-3059-port to master * commit '0e21e5b4d988d43040573a4fc71547829c815ec0': RED-3059: Fixed inital expansion overlaps	2021-12-10 11:37:24 +01:00
deiflaender	0e21e5b4d9	RED-3059: Fixed inital expansion overlaps	2021-12-10 11:32:53 +01:00
Dominique Eiflaender	29a66ee5cf	Pull request #290 : RED-2439: Added endpoint to calculate textBefore and textAfter for manual redactions Merge in RED/redaction-service from RED-2439 to master * commit '6950c00711dc113bb012a9cd3cc26c5c276409a0': RED-2439: Added endpoint to calculate textBefore and textAfter for manual redactions	2021-12-09 16:19:24 +01:00
deiflaender	6950c00711	RED-2439: Added endpoint to calculate textBefore and textAfter for manual redactions	2021-12-09 16:12:02 +01:00
Philipp Schramm	fa4b2263bc	Pull request #288 : RED-2795 Bugfix with inconsistent types for recategorizing Images in specific cases Merge in RED/redaction-service from bugfix/RED-2795 to master * commit '415df2b3658b4525120a15fb0f341164a73bb3d4': RED-2795 Bugfix with inconsistent types for recategorizing Images in specific cases	2021-12-07 11:15:30 +01:00
Philipp Schramm	415df2b365	RED-2795 Bugfix with inconsistent types for recategorizing Images in specific cases	2021-12-07 11:10:53 +01:00