Merge branch 'RED-9103' into 'release/0.78.x'

RED-9103: Fixed save of document viewer file

See merge request fforesight/layout-parser!144

.gitignore

@@ -38,3 +38,7 @@ build/
 
 ### VS Code ###
 .vscode/
+gradlew.bat
+gradlew
+gradle.properties
+gradle/

.gitlab-ci.yml

@@ -1,4 +1,23 @@
 include:
   - project: 'gitlab/gitlab'
     ref: 'main'
-    file: 'ci-templates/gradle_java.yml'
+    file: 'ci-templates/gradle_java.yml'
+
+deploy:
+  stage: deploy
+  tags:
+    - dind
+  script:
+    - echo "Building with gradle version ${BUILDVERSION}"
+    - gradle -Pversion=${BUILDVERSION} publish
+    - gradle bootBuildImage --cleanCache --publishImage -PbuildbootDockerHostNetwork=true -Pversion=${BUILDVERSION}
+    - echo "BUILDVERSION=$BUILDVERSION" >> version.env
+  artifacts:
+    reports:
+      dotenv: version.env
+  rules:
+    - if: $CI_COMMIT_BRANCH == $CI_DEFAULT_BRANCH
+    - if: $CI_COMMIT_BRANCH =~ /^release/
+    - if: $CI_COMMIT_TAG
+pmd:
+  allow_failure: true

README.md

@@ -1 +1,89 @@
+# PDF Layout Parser Micro-Service: layout-parser  
 
+## Introduction
+The layout-parser micro-service is a powerful tool designed to efficiently extract structured information from PDF documents. Written in Java and utilizing Spring Boot 3, Apache PDFBox, and RabbitMQ, this micro-service excels at parsing PDFs and organizing their content into a meaningful and coherent layout structure. Notably, the layout-parser micro-service distinguishes itself by relying solely on advanced algorithms, rather than machine learning techniques.
+
+### Key Steps in the PDF Layout Parsing Process:
+
+* **Text Position Extraction:**  
+The micro-service leverages Apache PDFBox to extract precise text positions for each individual character within the PDF document.
+
+* **Word Segmentation and Text Block Formation:**  
+Employing an array of diverse algorithms, the micro-service initially identifies and segments words, creating distinct text blocks.
+
+* **Text Block Classification:**  
+The segmented text blocks are then subjected to classification algorithms. These algorithms categorize the text blocks based on their content and visual properties, distinguishing between sections, subsections, headlines, paragraphs, images, tables, table cells, headers, and footers.
+
+* **Layout Coherence Establishment:**  
+The classified text blocks are subsequently orchestrated into a cohesive layout structure. This process involves arranging sections, subsections, paragraphs, images, and other elements in a logical and structured manner.
+
+* **Output Generation in Various Formats:**  
+Once the layout structure is established, the micro-service generates output in multiple formats. These formats are designed for seamless integration with downstream micro-services. The supported formats include JSON, XML, and others, ensuring flexibility in downstream data consumption.
+
+### Optional Enhancements:
+
+* **ML-Based Table Extraction:**  
+For enhanced results, users have the option to incorporate machine learning-based table extraction. This feature can be activated by providing ML-generated results as a JSON file, which are then integrated seamlessly into the layout structure.
+
+* **Image Classification using ML:**  
+Additionally, for more accurate image classification, users can optionally feed ML-generated image classification results into the micro-service. Similar to the table extraction option, the micro-service processes the pre-parsed results in JSON format, thus optimizing the accuracy of image content identification.
+
+In conclusion, the layout-parser micro-service is a versatile PDF layout parsing solution crafted entirely around advanced algorithms, without reliance on machine learning. It proficiently extracts text positions, segments content into meaningful blocks, classifies these blocks, arranges them coherently, and outputs structured data for downstream micro-services. Optional integration with ML-generated table extractions and image classifications further enhances its capabilities.
+
+
+
+
+
+
+## Installation
+
+### Prerequisites
+
+Before building and using the layout-parser micro-service, please ensure you have the following software and tools installed:
+
+Java Development Kit (JDK) 17 or later
+Gradle build tool (preinstalled)
+Build and Test
+To build and test the micro-service, follow these steps:
+
+### Clone the Repository:
+
+bash
+```
+git clone ssh://git@git.knecon.com:22222/fforesight/layout-parser.git
+cd layout-parser
+```
+### Build the Project:
+Use the following command to build the project using Gradle:
+
+```
+gradle clean build
+```
+### Run Tests:
+Run the test suite using the following command:
+```
+gradle test
+```
+## Building a Custom Docker Image
+To create a custom Docker image for the layout-parser micro-service, execute the provided script:
+
+### Ensure Docker is Installed:
+Ensure that Docker is installed and running on your system.
+
+### Run the Image Building Script:
+Execute the publish-custom-image script in the project directory:
+
+```
+./publish-custom-image
+```
+## Publishing to Internal Maven Repository
+To publish the layout-parser micro-service to your internal Maven repository, execute the following command:
+
+```
+gradle -Pversion=buildVersion publish
+```
+Replace buildVersion with the desired version number.
+
+## Additional Notes
+Make sure to configure any necessary application properties before deploying the micro-service.
+For advanced usage and configurations, refer to Kilian or Dom or preferably the source code.

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

@@ -1,24 +1,16 @@
 plugins {
-    java
+    `java-library`
+    `maven-publish`
     pmd
     checkstyle
     jacoco
 }
 
 group = "com.knecon.fforesight"
-version = "0.1-SNAPSHOT"
 
 java.sourceCompatibility = JavaVersion.VERSION_17
 java.targetCompatibility = JavaVersion.VERSION_17
 
-tasks.jacocoTestReport {
-    reports {
-        xml.required.set(false)
-        csv.required.set(false)
-        html.outputLocation.set(layout.buildDirectory.dir("jacocoHtml"))
-    }
-}
-
 tasks.pmdMain {
     pmd.ruleSetFiles = files("${rootDir}/config/pmd/pmd.xml")
 }
@@ -29,6 +21,9 @@ tasks.pmdTest {
 
 tasks.named<Test>("test") {
     useJUnitPlatform()
+    reports {
+        junitXml.outputLocation.set(layout.buildDirectory.dir("reports/junit"))
+    }
 }
 
 tasks.test {
@@ -40,14 +35,38 @@ tasks.jacocoTestReport {
     reports {
         xml.required.set(true)
         csv.required.set(false)
+        html.outputLocation.set(layout.buildDirectory.dir("jacocoHtml"))
     }
 }
 
+allprojects {
+    publishing {
+        publications {
+            create<MavenPublication>(name) {
+                from(components["java"])
+            }
+        }
+        repositories {
+            maven {
+                url = uri("https://nexus.knecon.com/repository/red-platform-releases/")
+                credentials {
+                    username = providers.gradleProperty("mavenUser").getOrNull();
+                    password = providers.gradleProperty("mavenPassword").getOrNull();
+                }
+            }
+        }
+    }
+}
+
+java {
+    withJavadocJar()
+}
+
 repositories {
     mavenLocal()
     mavenCentral()
     maven {
-        url = uri("https://nexus.knecon.com/repository/gindev/");
+        url = uri("https://nexus.knecon.com/repository/gindev/")
         credentials {
             username = providers.gradleProperty("mavenUser").getOrNull();
             password = providers.gradleProperty("mavenPassword").getOrNull();

config/pmd/pmd.xml

@@ -12,7 +12,6 @@
         <exclude name="DataflowAnomalyAnalysis"/>
         <exclude name="MissingSerialVersionUID"/>
         <exclude name="NullAssignment"/>
-        <exclude name="BeanMembersShouldSerialize"/>
         <exclude name="AvoidLiteralsInIfCondition"/>
         <exclude name="AvoidDuplicateLiterals"/>
         <exclude name="AvoidFieldNameMatchingMethodName"/>

config/pmd/test_pmd.xml

@@ -13,11 +13,11 @@
         <exclude name="DataflowAnomalyAnalysis"/>
         <exclude name="MissingSerialVersionUID"/>
         <exclude name="NullAssignment"/>
-        <exclude name="BeanMembersShouldSerialize"/>
         <exclude name="AvoidLiteralsInIfCondition"/>
         <exclude name="AvoidDuplicateLiterals"/>
         <exclude name="AvoidFieldNameMatchingMethodName"/>
         <exclude name="AvoidFieldNameMatchingTypeName"/>
+        <exclude name="TestClassWithoutTestCases"/>
     </rule>
 
 </ruleset>

gradle.properties.kts

@@ -1 +1 @@
-version = 0.1.0
+version = 0.1-SNAPSHOT

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

@@ -1,6 +1,10 @@
 plugins {
     id("com.knecon.fforesight.java-conventions")
-    id("io.freefair.lombok") version "8.1.0"
+    id("io.freefair.lombok") version "8.2.2"
 }
 
 description = "layoutparser-service-internal-api"
+
+dependencies {
+    implementation("io.swagger.core.v3:swagger-annotations:2.2.15")
+}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/DocumentData.java

@@ -1,5 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -10,11 +11,16 @@ import lombok.experimental.FieldDefaults;
 @Builder
 @AllArgsConstructor
 @FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+@Schema(description = "Object containing the complete document layout parsing information. It is split into 4 categories, structure, text, positions and pages: " + "The document tree structure of SemanticNodes such as Section, Paragraph, Headline, etc. " + "The text, which is stored as separate blocks of data. " + "The text positions, which are also stored as separate blocks. The Blocks are equal to the text blocks in length and order. " + "The page information.")
 public class DocumentData {
 
+    @Schema(description = "Contains information about the document's pages.")
     DocumentPage[] documentPages;
+    @Schema(description = "Contains information about the document's text.")
     DocumentTextData[] documentTextData;
+    @Schema(description = "Contains information about the document's text positions.")
     DocumentPositionData[] documentPositions;
+    @Schema(description = "Contains information about the document's semantic structure.")
     DocumentStructure documentStructure;
 
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/DocumentPage.java

@@ -1,5 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -12,11 +13,16 @@ import lombok.experimental.FieldDefaults;
 @NoArgsConstructor
 @AllArgsConstructor
 @FieldDefaults(level = AccessLevel.PRIVATE)
+@Schema(description = "Object containing information about the document's pages.")
 public class DocumentPage {
 
+    @Schema(description = "The page number, starting with 1.")
     int number;
+    @Schema(description = "The page height in PDF user units.", example = "792")
     int height;
+    @Schema(description = "The page width in PDF user units.", example = "694")
     int width;
+    @Schema(description = "The page rotation as specified by the PDF.", example = "90", allowableValues = {"0", "90", "180", "270"})
     int rotation;
 
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/DocumentPositionData.java

@@ -1,5 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -12,10 +13,14 @@ import lombok.experimental.FieldDefaults;
 @NoArgsConstructor
 @AllArgsConstructor
 @FieldDefaults(level = AccessLevel.PRIVATE)
+@Schema(description = "Object containing text positional information of a specific text block. A document is split into multiple text blocks, which are supposed to be read in order. Every text block can only occur on a single page.")
 public class DocumentPositionData {
 
+    @Schema(description = "Identifier of the text block.")
     Long id;
+    @Schema(description = "For each string coordinate in the search text of the text block, the array contains an entry relating the string coordinate to the position coordinate. This is required due to the text and position coordinates not being equal.")
     int[] stringIdxToPositionIdx;
+    @Schema(description = "The bounding box for each glyph as a rectangle. This matrix is of size (n,4), where n is the number of glyphs in the text block. The second dimension specifies the rectangle with the value x, y, width, height, with x, y specifying the lower left corner. In order to access this information, the stringIdxToPositionIdx array must be used to transform the coordinates.")
     float[][] positions;
 
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/DocumentStructure.java

@@ -1,9 +1,12 @@
 package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
+import java.awt.geom.Rectangle2D;
+import java.util.Arrays;
 import java.util.List;
 import java.util.Map;
 import java.util.stream.Stream;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -16,11 +19,50 @@ import lombok.experimental.FieldDefaults;
 @NoArgsConstructor
 @AllArgsConstructor
 @FieldDefaults(level = AccessLevel.PRIVATE)
+@Schema(description = "Object containing information about the parsed tree structure of the SemanticNodes, such as Section, Paragraph, Headline etc inside of the document.")
 public class DocumentStructure {
 
+    @Schema(description = "The root EntryData represents the Document.")
     EntryData root;
 
 
+    @Schema(description = "Object containing the extra field names, a table has in its properties field.")
+    public static class TableProperties {
+
+        public static final String NUMBER_OF_ROWS = "numberOfRows";
+        public static final String NUMBER_OF_COLS = "numberOfCols";
+
+    }
+
+    @Schema(description = "Object containing the extra field names, an Image has in its properties field.")
+    public static class ImageProperties {
+
+        public static final String TRANSPARENT = "transparent";
+        public static final String IMAGE_TYPE = "imageType";
+        public static final String POSITION = "position";
+        public static final String ID = "id";
+
+    }
+
+    @Schema(description = "Object containing the extra field names, a table cell has in its properties field.")
+    public static class TableCellProperties {
+
+        public static final String B_BOX = "bBox";
+        public static final String ROW = "row";
+        public static final String COL = "col";
+        public static final String HEADER = "header";
+
+    }
+
+    public static final String RECTANGLE_DELIMITER = ";";
+
+    public static Rectangle2D parseRectangle2D(String bBox) {
+
+        List<Float> floats = Arrays.stream(bBox.split(RECTANGLE_DELIMITER)).map(Float::parseFloat).toList();
+        return new Rectangle2D.Float(floats.get(0), floats.get(1), floats.get(2), floats.get(3));
+    }
+
+
     public EntryData get(List<Integer> tocId) {
 
         if (tocId.isEmpty()) {
@@ -57,15 +99,23 @@ public class DocumentStructure {
     @NoArgsConstructor
     @AllArgsConstructor
     @FieldDefaults(level = AccessLevel.PRIVATE)
+    @Schema(description = "Object containing information of a SemanticNode and also structuring the layout with children.")
     public static class EntryData {
 
+        @Schema(description = "Type of the semantic node.", allowableValues = {"DOCUMENT", "SECTION", "PARAGRAPH", "HEADLINE", "TABLE", "TABLE_CELL", "HEADER", "FOOTER", "IMAGE"})
         NodeType type;
+        @Schema(description = "Specifies the position in the parsed tree structure.", example = "[1, 0, 2]")
         int[] treeId;
+        @Schema(description = "Specifies the text block IDs associated with this semantic node. The value should be joined with the DocumentTextData/DocumentPositionData. Is empty, if no text block is directly associated with this semantic node. Only Paragraph, Headline, Header or Footer is directly associated with a text block.", example = "[1]")
         Long[] atomicBlockIds;
+        @Schema(description = "Specifies the pages this semantic node appears on. The value should be joined with the PageData.", example = "[1, 2, 3]")
         Long[] pageNumbers;
+        @Schema(description = "Some semantic nodes have additional information, this information is stored in this Map. The extra fields are specified by the Properties subclasses.", example = "For a Table: {\"numberOfRows\": 3, \"numberOfCols\": 4}")
         Map<String, String> properties;
+        @Schema(description = "All child Entries of this Entry.", example = "[1, 2, 3]")
         List<EntryData> children;
 
+
         @Override
         public String toString() {
 

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/DocumentTextData.java

@@ -1,7 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
 
-
-
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AccessLevel;
 import lombok.AllArgsConstructor;
 import lombok.Builder;
@@ -14,14 +13,22 @@ import lombok.experimental.FieldDefaults;
 @NoArgsConstructor
 @AllArgsConstructor
 @FieldDefaults(level = AccessLevel.PRIVATE)
+@Schema(description = "Object containing text information of a specific text block. A document is split into multiple text blocks, which are supposed to be read in order. Every text block can only occur on a single page.")
 public class DocumentTextData {
 
+    @Schema(description = "Identifier of the text block.")
     Long id;
+    @Schema(description = "The page the text block occurs on.")
     Long page;
+    @Schema(description = "The text the text block.")
     String searchText;
+    @Schema(description = "Each text block is assigned a number on a page, starting from 0.")
     int numberOnPage;
+    @Schema(description = "The text blocks are ordered, this number represents the start of the text block as a string offset.")
     int start;
+    @Schema(description = "The text blocks are ordered, this number represents the end of the text block as a string offset.")
     int end;
+    @Schema(description = "The line breaks in the text of this semantic node in string offsets. They are exclusive end. At the end of each semantic node there is an implicit linebreak.", example = "[5, 10]")
     int[] lineBreaks;
 
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/SimplifiedSectionText.java

@@ -0,0 +1,21 @@
+package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.NoArgsConstructor;
+
+@Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
+@Schema(description = "Object containing a simplified version, which contains almost exclusively text, of the document structure Section class.")
+public class SimplifiedSectionText {
+
+    @Schema(description = "The number of this Section. This is used to map the simplified section text back to the original Section.")
+    private int sectionNumber;
+    @Schema(description = "The text in this Section.")
+    private String text;
+
+}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/redaction/SimplifiedText.java

@@ -0,0 +1,24 @@
+package com.knecon.fforesight.service.layoutparser.internal.api.data.redaction;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import io.swagger.v3.oas.annotations.media.Schema;
+import lombok.AllArgsConstructor;
+import lombok.Builder;
+import lombok.Data;
+import lombok.NoArgsConstructor;
+
+@Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
+@Schema(description = "Object containing a simplified version, which contains almost exclusively text, of the document structure.")
+public class SimplifiedText {
+
+    @Schema(description = "Number of pages in the entire document.")
+    private int numberOfPages;
+    @Schema(description = "A List of simplified Sections, which contains almost exclusively text.")
+    private List<SimplifiedSectionText> sectionTexts = new ArrayList<>();
+
+}

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/ParagraphData.java

@@ -2,20 +2,29 @@ package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
 
 import java.util.List;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.Builder;
 import lombok.Data;
 
 @Data
 @Builder
+@Schema(description = "Object containing information about a Paragraph/Headline/Header/Footer.")
 public class ParagraphData {
 
+    @Schema(description = "The text of this Semantic Node, without any linebreaks.", example = "This is some text.")
     private String text;
+    @Schema(description = "A list of text ranges in string offsets. Every character in any of the ranges is bold.", example = "[0, 15]")
     List<Range> boldTextBoundaries;
+    @Schema(description = "A list of text ranges in string offsets. Every character in any of the ranges is italic.", example = "[0, 15]")
     List<Range> italicTextBoundaries;
+    @Schema(description = "The line breaks in the text of this semantic node in string offsets. They are exclusive end. At the end of each semantic node there is an implicit linebreak.", example = "[5, 10]")
     List<Integer> linebreaks;
+    @Schema(description = "The classification of this Paragraph.", allowableValues = "{paragraph, headline, header, footer}")
     private String classification;
 
+    @Schema(description = "Describes the text orientation of this semantic node. Any semantic node only has a single text orientation.", allowableValues = "{ZERO, QUARTER_CIRCLE, HALF_CIRCLE, THREE_QUARTER_CIRCLE}")
     private String orientation;
+    @Schema(description = "Describes the text direction in degrees of this semantic node. Any semantic node only has a single text direction.", minimum = "0", maximum = "359")
     private int textDirection;
 
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/Range.java

@@ -1,5 +1,8 @@
 package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
 
+import io.swagger.v3.oas.annotations.media.Schema;
+
+@Schema(description = "Object specifying the start and end offsets of a text range in string offsets.")
 public record Range(int start, int end) {
 
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/ResearchDocumentData.java

@@ -2,6 +2,7 @@ package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
 
 import java.util.List;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AllArgsConstructor;
 import lombok.Builder;
 import lombok.Data;
@@ -9,8 +10,12 @@ import lombok.Data;
 @Builder
 @Data
 @AllArgsConstructor
+@Schema(description = "Object containing a simplified version of the document structure. This simplified form only knows Paragraphs and Tables. The Paragraph Objects might be a Paragraph, Headline, Header or Footer.")
 public class ResearchDocumentData {
 
+    @Schema(description = "File name of the original uploaded file.")
     String originalFile;
+    @Schema(description = "A List of all paragraphs/headline or table objects, that have been parsed in this document.")
     List<StructureObject> structureObjects;
+
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/RowData.java

@@ -2,14 +2,19 @@ package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
 
 import java.util.List;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AllArgsConstructor;
 import lombok.Data;
 
 @Data
 @AllArgsConstructor
+@Schema(description = "Object containing information about a Table Row.")
 public class RowData {
 
+    @Schema(description = "Boolean indicating whether this table row is classified as a header row.")
     boolean header;
+    @Schema(description = "A list of Objects containing information about the text in each cell of this row.")
     List<ParagraphData> cellText;
+    @Schema(description = "The bounding box of this StructureObject. Is always exactly 4 values representing x, y, w, h, where x, y specify the lower left corner.")
     float[] bBox;
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/StructureObject.java

@@ -1,5 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AllArgsConstructor;
 import lombok.Builder;
 import lombok.Data;
@@ -7,13 +8,20 @@ import lombok.Data;
 @Data
 @Builder
 @AllArgsConstructor
+@Schema(description = "Object containing information about either a Paragraph/Headline/Header/Footer or a Table.")
 public class StructureObject {
 
+    @Schema(description = "The ID of this StructureObject.")
     Integer structureObjectNumber;
+    @Schema(description = "This value indicates the start of the string offsets in this Object, with respect to the reading order.")
     int page;
+    @Schema(description = "This stringOffset indicates the start of the string offsets in this Object, with respect to the reading order of the entire document. It is equal to the previous' StructureObject stringOffset + its length.")
     int stringOffset;
+    @Schema(description = "The bounding box of this StructureObject. Is always exactly 4 values representing x, y, w, h, where x, y specify the lower left corner.", example = "[100, 100, 50, 50]")
     float[] boundingBox;
+    @Schema(description = "Object containing information about a Paragraph/Headline/Header/Footer. Either this or table is null.")
     ParagraphData paragraph;
+    @Schema(description = "Object containing information about a Table. Either this or paragraph is null.")
     TableData table;
 
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/data/taas/TableData.java

@@ -2,15 +2,20 @@ package com.knecon.fforesight.service.layoutparser.internal.api.data.taas;
 
 import java.util.List;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.AllArgsConstructor;
 import lombok.Data;
 
 @Data
 @AllArgsConstructor
+@Schema(description = "Object containing information about a Table.")
 public class TableData {
 
+    @Schema(description = "A list of Objects containing information about all rows in this table.")
     List<RowData> rowData;
+    @Schema(description = "Number of columns in this table.")
     Integer numberOfCols;
+    @Schema(description = "Number of rows in this table.")
     Integer numberOfRows;
 
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/queue/LayoutParsingFinishedEvent.java

@@ -2,9 +2,19 @@ package com.knecon.fforesight.service.layoutparser.internal.api.queue;
 
 import java.util.Map;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.Builder;
 
 @Builder
-public record LayoutParsingFinishedEvent(Map<String, String> identifier, long duration, int numberOfPages, String message) {
+@Schema(description = "Object containing information about the layout parsing.")
+public record LayoutParsingFinishedEvent(
+        @Schema(description = "General purpose identifier. It is returned exactly the same way it is inserted with the LayoutParsingRequest.")
+        Map<String, String> identifier,//
+        @Schema(description = "The duration of a single layout parsing in ms.")
+        long duration,//
+        @Schema(description = "The number of pages of the parsed document.")
+        int numberOfPages,//
+        @Schema(description = "A general message. It contains some information useful for a developer, like the paths where the files are stored. Not meant to be machine readable.")
+        String message) {
 
 }

layoutparser-service/layoutparser-service-internal-api/src/main/java/com/knecon/fforesight/service/layoutparser/internal/api/queue/LayoutParsingRequest.java

@@ -3,21 +3,39 @@ package com.knecon.fforesight.service.layoutparser.internal.api.queue;
 import java.util.Map;
 import java.util.Optional;
 
+import io.swagger.v3.oas.annotations.media.Schema;
 import lombok.Builder;
-
+import lombok.NonNull;
 
 @Builder
+@Schema(description = "Object containing all storage paths the service needs to know.")
 public record LayoutParsingRequest(
-        LayoutParsingType layoutParsingType,
+        @Schema(description = "Enum specifying the type of layout parsing to be performed.", allowableValues = "{RedactManager, DocuMine, TAAS}")//
+        @NonNull LayoutParsingType layoutParsingType,
+
+        @Schema(description = "General purpose identifiers. They are not changed by the service at all and are returned as is in the response queue.")//
         Map<String, String> identifier,
-        String originFileStorageId,
-        Optional<String> tablesFileStorageId,
-        Optional<String> imagesFileStorageId,
-        String structureFileStorageId,
-        String researchDocumentStorageId,
-        String textBlockFileStorageId,
-        String positionBlockFileStorageId,
-        String pageFileStorageId,
-        String sectionGridStorageId) {
+
+        @Schema(description = "Path to the original PDF file.")//
+        @NonNull String originFileStorageId,//
+        @Schema(description = "Optional Path to the table extraction file.")//
+        Optional<String> tablesFileStorageId,//
+        @Schema(description = "Optional Path to the image classification file.")//
+        Optional<String> imagesFileStorageId,//
+
+        @Schema(description = "Path where the Document Structure File will be stored.")//
+        @NonNull String structureFileStorageId,//
+        @Schema(description = "Path where the Research Data File will be stored.")//
+        String researchDocumentStorageId,//
+        @Schema(description = "Path where the Document Text File will be stored.")//
+        @NonNull String textBlockFileStorageId,//
+        @Schema(description = "Path where the Document Positions File will be stored.")//
+        @NonNull String positionBlockFileStorageId,//
+        @Schema(description = "Path where the Document Pages File will be stored.")//
+        @NonNull String pageFileStorageId,//
+        @Schema(description = "Path where the Simplified Text File will be stored.")//
+        @NonNull String simplifiedTextStorageId,//
+        @Schema(description = "Path where the Viewer Document PDF will be stored.")//
+        @NonNull String viewerDocumentStorageId) {
 
 }

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

@@ -1,21 +1,26 @@
 plugins {
     id("com.knecon.fforesight.java-conventions")
-    id("io.freefair.lombok") version "8.1.0"
+    id("io.freefair.lombok") version "8.2.2"
 }
 
+description = "layoutparser-service-processor"
+
+val jacksonVersion = "2.15.2"
+val pdfBoxVersion = "3.0.0"
+
 dependencies {
     implementation(project(":layoutparser-service-internal-api"))
 
-    implementation("com.iqser.red.service:persistence-service-shared-api-v1:2.36.0")
+    implementation("com.iqser.red.service:persistence-service-shared-api-v1:2.144.0") {
+        exclude("org.springframework.boot", "spring-boot-starter-security")
+        exclude("org.springframework.boot", "spring-boot-starter-validation")
+    }
     implementation("com.knecon.fforesight:tenant-commons:0.10.0")
-    implementation("com.iqser.red.commons:storage-commons:2.1.0")
+    implementation("com.iqser.red.commons:storage-commons:2.40.0")
 
-    implementation("org.apache.pdfbox:pdfbox:3.0.0-alpha2")
-    implementation("org.apache.pdfbox:pdfbox-tools:3.0.0-alpha2")
-    implementation("com.fasterxml.jackson.module:jackson-module-afterburner:2.15.0-rc2")
-    implementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:2.15.0-rc2")
-    implementation("org.springframework.boot:spring-boot-starter-web:3.0.6")
-    implementation("org.springframework.boot:spring-boot-starter-amqp:3.0.6")
+    implementation("org.apache.pdfbox:pdfbox:${pdfBoxVersion}")
+    implementation("org.apache.pdfbox:pdfbox-tools:${pdfBoxVersion}")
+    implementation("com.fasterxml.jackson.module:jackson-module-afterburner:${jacksonVersion}")
+    implementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:${jacksonVersion}")
+    implementation("org.springframework.boot:spring-boot-starter-web:3.1.3")
 }
-
-description = "layoutparser-service-processor"

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

@@ -2,31 +2,57 @@ package com.knecon.fforesight.service.layoutparser.processor;
 
 import static java.lang.String.format;
 
+import java.awt.geom.Rectangle2D;
+import java.io.ByteArrayOutputStream;
 import java.io.IOException;
+import java.util.ArrayList;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
 
 import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.common.PDRectangle;
+import org.apache.pdfbox.pdmodel.documentinterchange.markedcontent.PDMarkedContent;
 import org.springframework.stereotype.Service;
 
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingFinishedEvent;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingRequest;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
+import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
+import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
+import com.knecon.fforesight.service.layoutparser.processor.model.image.ClassifiedImage;
+import com.knecon.fforesight.service.layoutparser.processor.model.table.CleanRulings;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.adapter.CvTableParsingAdapter;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.adapter.ImageServiceResponseAdapter;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.image.ImageServiceResponse;
+import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableCells;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
-import com.knecon.fforesight.service.layoutparser.processor.services.factory.DocumentGraphFactory;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
-import com.knecon.fforesight.service.layoutparser.processor.services.mapper.DocumentDataMapper;
-import com.knecon.fforesight.service.layoutparser.processor.services.mapper.TaasDocumentDataMapper;
-import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
-import com.knecon.fforesight.service.layoutparser.processor.services.PdfParsingService;
-import com.knecon.fforesight.service.layoutparser.processor.services.SectionGridCreatorService;
+import com.knecon.fforesight.service.layoutparser.processor.services.BodyTextFrameService;
+import com.knecon.fforesight.service.layoutparser.processor.services.RulingCleaningService;
 import com.knecon.fforesight.service.layoutparser.processor.services.SectionsBuilderService;
+import com.knecon.fforesight.service.layoutparser.processor.services.SimplifiedSectionTextService;
+import com.knecon.fforesight.service.layoutparser.processor.services.TableExtractionService;
+import com.knecon.fforesight.service.layoutparser.processor.services.blockification.DocuMineBlockificationService;
+import com.knecon.fforesight.service.layoutparser.processor.services.blockification.RedactManagerBlockificationService;
+import com.knecon.fforesight.service.layoutparser.processor.services.blockification.TaasBlockificationService;
 import com.knecon.fforesight.service.layoutparser.processor.services.classification.DocuMineClassificationService;
 import com.knecon.fforesight.service.layoutparser.processor.services.classification.RedactManagerClassificationService;
 import com.knecon.fforesight.service.layoutparser.processor.services.classification.TaasClassificationService;
+import com.knecon.fforesight.service.layoutparser.processor.services.factory.DocumentGraphFactory;
+import com.knecon.fforesight.service.layoutparser.processor.services.mapper.DocumentDataMapper;
+import com.knecon.fforesight.service.layoutparser.processor.services.mapper.TaasDocumentDataMapper;
+import com.knecon.fforesight.service.layoutparser.processor.services.parsing.PDFLinesTextStripper;
+import com.knecon.fforesight.service.layoutparser.processor.services.visualization.ViewerDocumentService;
+import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 
 import lombok.RequiredArgsConstructor;
+import lombok.SneakyThrows;
 import lombok.extern.slf4j.Slf4j;
 
 @Slf4j
@@ -37,12 +63,18 @@ public class LayoutParsingPipeline {
     private final ImageServiceResponseAdapter imageServiceResponseAdapter;
     private final CvTableParsingAdapter cvTableParsingAdapter;
     private final LayoutParsingStorageService layoutParsingStorageService;
-    private final PdfParsingService pdfParsingService;
     private final SectionsBuilderService sectionsBuilderService;
-    private final SectionGridCreatorService sectionGridCreatorService;
     private final TaasClassificationService taasClassificationService;
     private final RedactManagerClassificationService redactManagerClassificationService;
     private final DocuMineClassificationService docuMineClassificationService;
+    private final SimplifiedSectionTextService simplifiedSectionTextService;
+    private final BodyTextFrameService bodyTextFrameService;
+    private final RulingCleaningService rulingCleaningService;
+    private final TableExtractionService tableExtractionService;
+    private final TaasBlockificationService taasBlockificationService;
+    private final DocuMineBlockificationService docuMineBlockificationService;
+    private final RedactManagerBlockificationService redactManagerBlockificationService;
+    private final ViewerDocumentService viewerDocumentService;
 
 
     public LayoutParsingFinishedEvent parseLayoutAndSaveFilesToStorage(LayoutParsingRequest layoutParsingRequest) throws IOException {
@@ -52,19 +84,26 @@ public class LayoutParsingPipeline {
         try (PDDocument originDocument = layoutParsingStorageService.getOriginFile(layoutParsingRequest.originFileStorageId())) {
             ImageServiceResponse imageServiceResponse = new ImageServiceResponse();
             if (layoutParsingRequest.imagesFileStorageId().isPresent()) {
-                imageServiceResponse = layoutParsingStorageService.getImagesFile(layoutParsingRequest.pageFileStorageId());
+                imageServiceResponse = layoutParsingStorageService.getImagesFile(layoutParsingRequest.imagesFileStorageId().get());
             }
 
             TableServiceResponse tableServiceResponse = new TableServiceResponse();
             if (layoutParsingRequest.tablesFileStorageId().isPresent()) {
-                tableServiceResponse = layoutParsingStorageService.getTablesFile(layoutParsingRequest.pageFileStorageId());
+                tableServiceResponse = layoutParsingStorageService.getTablesFile(layoutParsingRequest.tablesFileStorageId().get());
             }
 
-            Document documentGraph = parseLayout(layoutParsingRequest.layoutParsingType(), originDocument, imageServiceResponse, tableServiceResponse);
+            ClassificationDocument classificationDocument = parseLayout(layoutParsingRequest.layoutParsingType(), originDocument, imageServiceResponse, tableServiceResponse);
+            Document documentGraph = DocumentGraphFactory.buildDocumentGraph(classificationDocument);
+
             int numberOfPages = originDocument.getNumberOfPages();
 
-            layoutParsingStorageService.storeSectionGrid(layoutParsingRequest, sectionGridCreatorService.createSectionGrid(documentGraph));
             layoutParsingStorageService.storeDocumentData(layoutParsingRequest, DocumentDataMapper.toDocumentData(documentGraph));
+            layoutParsingStorageService.storeSimplifiedText(layoutParsingRequest, simplifiedSectionTextService.toSimplifiedText(documentGraph));
+
+            try (var out = new ByteArrayOutputStream()) {
+                viewerDocumentService.createViewerDocument(originDocument, documentGraph, out, false);
+                layoutParsingStorageService.storeViewerDocument(layoutParsingRequest, out);
+            }
 
             if (layoutParsingRequest.layoutParsingType().equals(LayoutParsingType.TAAS)) {
                 var researchDocumentData = TaasDocumentDataMapper.fromDocument(documentGraph);
@@ -75,25 +114,111 @@ public class LayoutParsingPipeline {
                     .identifier(layoutParsingRequest.identifier())
                     .numberOfPages(numberOfPages)
                     .duration(System.currentTimeMillis() - start)
-                    .message(format("Layout parsing is finished and files have been saved with Ids:\n Structure: %s\nText: %s\nPositions: %s\nPageData: %s",
+                    .message(format("""
+                                    Layout parsing has finished in %.02f s.
+                                    identifiers: %s
+                                    %s
+                                    Files have been saved with Ids:
+                                    Structure: %s
+                                    Text: %s
+                                    Positions: %s
+                                    PageData: %s
+                                    Simplified Text: %s
+                                    Viewer Doc: %s""",
+                            ((float) (System.currentTimeMillis() - start)) / 1000,
+                            layoutParsingRequest.identifier(),
+                            buildSemanticNodeCountMessage(numberOfPages, documentGraph.buildSemanticNodeCounts()),
                             layoutParsingRequest.structureFileStorageId(),
                             layoutParsingRequest.textBlockFileStorageId(),
                             layoutParsingRequest.positionBlockFileStorageId(),
-                            layoutParsingRequest.pageFileStorageId()))
+                            layoutParsingRequest.pageFileStorageId(),
+                            layoutParsingRequest.simplifiedTextStorageId(),
+                            layoutParsingRequest.viewerDocumentStorageId()))
                     .build();
         }
     }
 
 
-    public Document parseLayout(LayoutParsingType layoutParsingType,
-                                PDDocument originDocument,
-                                ImageServiceResponse imageServiceResponse,
-                                TableServiceResponse tableServiceResponse) {
+    private String buildSemanticNodeCountMessage(int numberOfPages, Map<NodeType, Long> semanticNodeCounts) {
 
-        ClassificationDocument classificationDocument = pdfParsingService.parseDocument(layoutParsingType,
-                originDocument,
-                cvTableParsingAdapter.buildCvParsedTablesPerPage(tableServiceResponse),
-                imageServiceResponseAdapter.buildClassifiedImagesPerPage(imageServiceResponse));
+        return String.format("%d pages with %d sections, %d headlines, %d paragraphs, %d tables with %d cells, %d headers, and %d footers parsed",
+                numberOfPages,
+                semanticNodeCounts.get(NodeType.SECTION) == null ? 0 : semanticNodeCounts.get(NodeType.SECTION),
+                semanticNodeCounts.get(NodeType.HEADLINE) == null ? 0 : semanticNodeCounts.get(NodeType.HEADLINE),
+                semanticNodeCounts.get(NodeType.PARAGRAPH) == null ? 0 : semanticNodeCounts.get(NodeType.PARAGRAPH),
+                semanticNodeCounts.get(NodeType.TABLE) == null ? 0 : semanticNodeCounts.get(NodeType.TABLE),
+                semanticNodeCounts.get(NodeType.TABLE_CELL) == null ? 0 : semanticNodeCounts.get(NodeType.TABLE_CELL),
+                semanticNodeCounts.get(NodeType.HEADER) == null ? 0 : semanticNodeCounts.get(NodeType.HEADER),
+                semanticNodeCounts.get(NodeType.FOOTER) == null ? 0 : semanticNodeCounts.get(NodeType.FOOTER));
+    }
+
+
+    @SneakyThrows
+    public ClassificationDocument parseLayout(LayoutParsingType layoutParsingType,
+                                              PDDocument originDocument,
+                                              ImageServiceResponse imageServiceResponse,
+                                              TableServiceResponse tableServiceResponse) {
+
+        Map<Integer, List<TableCells>> pdfTableCells = cvTableParsingAdapter.buildCvParsedTablesPerPage(tableServiceResponse);
+        Map<Integer, List<ClassifiedImage>> pdfImages = imageServiceResponseAdapter.buildClassifiedImagesPerPage(imageServiceResponse);
+
+        ClassificationDocument classificationDocument = new ClassificationDocument();
+        List<ClassificationPage> classificationPages = new ArrayList<>();
+
+        originDocument.setAllSecurityToBeRemoved(true);
+        long pageCount = originDocument.getNumberOfPages();
+
+        for (int pageNumber = 1; pageNumber <= pageCount; pageNumber++) {
+            classificationDocument.setPages(classificationPages);
+            PDFLinesTextStripper stripper = new PDFLinesTextStripper();
+            PDPage pdPage = originDocument.getPage(pageNumber - 1);
+            stripper.setPageNumber(pageNumber);
+            stripper.setStartPage(pageNumber);
+            stripper.setEndPage(pageNumber);
+            stripper.setPdpage(pdPage);
+            if (layoutParsingType.equals(LayoutParsingType.DOCUMINE)) {
+                stripper.setSortByPosition(true);
+            }
+            stripper.getText(originDocument);
+
+            PDRectangle pdr = pdPage.getMediaBox();
+
+            int rotation = pdPage.getRotation();
+            boolean isLandscape = pdr.getWidth() > pdr.getHeight() && (rotation == 0 || rotation == 180) || pdr.getHeight() > pdr.getWidth() && (rotation == 90 || rotation == 270);
+
+            PDRectangle cropbox = pdPage.getCropBox();
+            CleanRulings cleanRulings = rulingCleaningService.getCleanRulings(pdfTableCells.get(pageNumber), stripper.getRulings());
+
+            ClassificationPage classificationPage = switch (layoutParsingType) {
+                case REDACT_MANAGER -> redactManagerBlockificationService.blockify(stripper.getTextPositionSequences(), cleanRulings.getHorizontal(), cleanRulings.getVertical());
+                case TAAS -> taasBlockificationService.blockify(stripper.getTextPositionSequences(), cleanRulings.getHorizontal(), cleanRulings.getVertical());
+                case DOCUMINE -> docuMineBlockificationService.blockify(stripper.getTextPositionSequences(), cleanRulings.getHorizontal(), cleanRulings.getVertical());
+            };
+            classificationPage.setCleanRulings(cleanRulings);
+            classificationPage.setRotation(rotation);
+            classificationPage.setLandscape(isLandscape);
+            classificationPage.setPageNumber(pageNumber);
+            classificationPage.setPageWidth(cropbox.getWidth());
+            classificationPage.setPageHeight(cropbox.getHeight());
+
+            // MarkedContent needs to be converted at this point, otherwise it leads to GC Problems in Pdfbox.
+            classificationPage.setMarkedContentBboxPerType(convertMarkedContents(stripper.getMarkedContents()));
+
+            // If images is ocr needs to be calculated before textBlocks are moved into tables, otherwise findOcr algorithm needs to be adopted.
+            if (pdfImages != null && pdfImages.containsKey(pageNumber)) {
+                classificationPage.setImages(pdfImages.get(pageNumber));
+                imageServiceResponseAdapter.findOcr(classificationPage);
+            }
+
+            tableExtractionService.extractTables(cleanRulings, classificationPage);
+
+            buildPageStatistics(classificationPage);
+            increaseDocumentStatistics(classificationPage, classificationDocument);
+
+            classificationPages.add(classificationPage);
+        }
+
+        bodyTextFrameService.setBodyTextFrames(classificationDocument, layoutParsingType);
 
         switch (layoutParsingType) {
             case TAAS -> taasClassificationService.classifyDocument(classificationDocument);
@@ -104,40 +229,47 @@ public class LayoutParsingPipeline {
         sectionsBuilderService.buildSections(classificationDocument);
         sectionsBuilderService.addImagesToSections(classificationDocument);
 
-        return DocumentGraphFactory.buildDocumentGraph(classificationDocument);
+        return classificationDocument;
     }
 
 
-    public Document parseLayoutWithTimer(LayoutParsingType layoutParsingType,
-                                         PDDocument originDocument,
-                                         ImageServiceResponse imageServiceResponse,
-                                         TableServiceResponse tableServiceResponse) {
+    private Map<String, List<Rectangle2D>> convertMarkedContents(List<PDMarkedContent> pdMarkedContents) {
 
-        long start = System.currentTimeMillis();
+        Map<String, List<Rectangle2D>> markedContentBboxes = new HashMap<>();
+        markedContentBboxes.put(MarkedContentUtils.HEADER, MarkedContentUtils.getMarkedContentBboxPerLine(pdMarkedContents, MarkedContentUtils.HEADER));
+        markedContentBboxes.put(MarkedContentUtils.FOOTER, MarkedContentUtils.getMarkedContentBboxPerLine(pdMarkedContents, MarkedContentUtils.FOOTER));
+        return markedContentBboxes;
+    }
 
-        ClassificationDocument classificationDocument = pdfParsingService.parseDocument(layoutParsingType,
-                originDocument,
-                cvTableParsingAdapter.buildCvParsedTablesPerPage(tableServiceResponse),
-                imageServiceResponseAdapter.buildClassifiedImagesPerPage(imageServiceResponse));
 
-        System.out.printf("parsed %d ms", System.currentTimeMillis() - start);
+    private void increaseDocumentStatistics(ClassificationPage classificationPage, ClassificationDocument document) {
 
-        start = System.currentTimeMillis();
-        switch (layoutParsingType) {
-            case TAAS -> taasClassificationService.classifyDocument(classificationDocument);
-            case DOCUMINE -> docuMineClassificationService.classifyDocument(classificationDocument);
-            case REDACT_MANAGER -> redactManagerClassificationService.classifyDocument(classificationDocument);
+       if (!classificationPage.isLandscape()) {
+        document.getFontSizeCounter().addAll(classificationPage.getFontSizeCounter().getCountPerValue());
+      }
+        document.getFontCounter().addAll(classificationPage.getFontCounter().getCountPerValue());
+        document.getTextHeightCounter().addAll(classificationPage.getTextHeightCounter().getCountPerValue());
+        document.getFontStyleCounter().addAll(classificationPage.getFontStyleCounter().getCountPerValue());
+    }
+
+
+    private void buildPageStatistics(ClassificationPage classificationPage) {
+
+        // Collect all statistics for the classificationPage, except from blocks inside tables, as tables will always be added to BodyTextFrame.
+        for (AbstractPageBlock textBlock : classificationPage.getTextBlocks()) {
+            if (textBlock instanceof TextPageBlock) {
+                if (((TextPageBlock) textBlock).getSequences() == null) {
+                    continue;
+                }
+                for (TextPositionSequence word : ((TextPageBlock) textBlock).getSequences()) {
+                    classificationPage.getTextHeightCounter().add(word.getTextHeight());
+                    classificationPage.getFontCounter().add(word.getFont());
+                    classificationPage.getFontSizeCounter().add(word.getFontSize());
+                    classificationPage.getFontStyleCounter().add(word.getFontStyle());
+                }
+            }
         }
-        System.out.printf(", classified %d ms", System.currentTimeMillis() - start);
 
-        start = System.currentTimeMillis();
-        sectionsBuilderService.buildSections(classificationDocument);
-        System.out.printf(", sections built %d ms", System.currentTimeMillis() - start);
-
-        start = System.currentTimeMillis();
-        Document document = DocumentGraphFactory.buildDocumentGraph(classificationDocument);
-        System.out.printf(", graph constructed %d ms", System.currentTimeMillis() - start);
-        return document;
     }
 
 }

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

@@ -1,25 +1,25 @@
 package com.knecon.fforesight.service.layoutparser.processor;
 
+import java.io.ByteArrayInputStream;
+import java.io.ByteArrayOutputStream;
 import java.io.File;
 import java.io.FileOutputStream;
 import java.io.IOException;
 import java.io.InputStream;
 import java.nio.file.Files;
+import java.nio.file.Path;
+import java.nio.file.Paths;
+import java.nio.file.StandardOpenOption;
 
 import org.apache.commons.io.IOUtils;
 import org.apache.pdfbox.Loader;
-import org.apache.pdfbox.io.MemoryUsageSetting;
 import org.apache.pdfbox.pdmodel.PDDocument;
 import org.springframework.stereotype.Service;
 
 import com.fasterxml.jackson.databind.ObjectMapper;
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.section.SectionGrid;
 import com.iqser.red.storage.commons.service.StorageService;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPositionData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentTextData;
 import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentData;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructure;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentPage;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.SimplifiedText;
 import com.knecon.fforesight.service.layoutparser.internal.api.data.taas.ResearchDocumentData;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingRequest;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.image.ImageServiceResponse;
@@ -27,6 +27,7 @@ import com.knecon.fforesight.service.layoutparser.processor.python_api.model.tab
 import com.knecon.fforesight.tenantcommons.TenantContext;
 
 import lombok.RequiredArgsConstructor;
+import lombok.SneakyThrows;
 import lombok.extern.slf4j.Slf4j;
 
 @Slf4j
@@ -40,31 +41,35 @@ public class LayoutParsingStorageService {
 
     public PDDocument getOriginFile(String storageId) throws IOException {
 
-        try (var originDocumentInputStream = storageService.getObject(TenantContext.getTenantId(), storageId).getInputStream()) {
+        try (var originDocumentInputStream = getObject(storageId)) {
             File tempFile = createTempFile("document", ".pdf");
             try (var tempFileOutputStream = new FileOutputStream(tempFile)) {
                 IOUtils.copy(originDocumentInputStream, tempFileOutputStream);
+                originDocumentInputStream.close();
             }
-            return Loader.loadPDF(tempFile, MemoryUsageSetting.setupMixed(67108864L));
+            return Loader.loadPDF(tempFile);
         }
     }
 
 
     public ImageServiceResponse getImagesFile(String storageId) throws IOException {
 
-        try (InputStream inputStream = storageService.getObject(TenantContext.getTenantId(), storageId).getInputStream()) {
+        try (InputStream inputStream = getObject(storageId)) {
 
-            return objectMapper.readValue(inputStream, ImageServiceResponse.class);
+            ImageServiceResponse imageServiceResponse = objectMapper.readValue(inputStream, ImageServiceResponse.class);
+            inputStream.close();
+            return imageServiceResponse;
         }
     }
 
 
     public TableServiceResponse getTablesFile(String storageId) throws IOException {
 
-        try (var tableClassificationStream = storageService.getObject(TenantContext.getTenantId(), storageId).getInputStream()) {
-
-            return objectMapper.readValue(tableClassificationStream, TableServiceResponse.class);
+        try (var tableClassificationStream = getObject(storageId)) {
 
+            TableServiceResponse tableServiceResponse = objectMapper.readValue(tableClassificationStream, TableServiceResponse.class);
+            tableClassificationStream.close();
+            return tableServiceResponse;
         }
     }
 
@@ -78,11 +83,6 @@ public class LayoutParsingStorageService {
     }
 
 
-    public void storeSectionGrid(LayoutParsingRequest layoutParsingRequest, SectionGrid sectionGrid) {
-
-        storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.sectionGridStorageId(), sectionGrid);
-    }
-
 
     public void storeResearchDocumentData(LayoutParsingRequest layoutParsingRequest, ResearchDocumentData researchDocumentData) {
 
@@ -90,26 +90,6 @@ public class LayoutParsingStorageService {
     }
 
 
-    public DocumentData readDocumentData(LayoutParsingRequest layoutParsingRequest) throws IOException {
-
-        DocumentPage[] documentPageData = storageService.readJSONObject(TenantContext.getTenantId(), layoutParsingRequest.pageFileStorageId(), DocumentPage[].class);
-        DocumentTextData[] documentTextDataBlockData = storageService.readJSONObject(TenantContext.getTenantId(),
-                layoutParsingRequest.textBlockFileStorageId(),
-                DocumentTextData[].class);
-        DocumentPositionData[] atomicPositionBlockData = storageService.readJSONObject(TenantContext.getTenantId(),
-                layoutParsingRequest.positionBlockFileStorageId(),
-                DocumentPositionData[].class);
-        DocumentStructure tableOfContentsData = storageService.readJSONObject(TenantContext.getTenantId(), layoutParsingRequest.structureFileStorageId(), DocumentStructure.class);
-
-        return DocumentData.builder()
-                .documentStructure(tableOfContentsData)
-                .documentPositions(atomicPositionBlockData)
-                .documentTextData(documentTextDataBlockData)
-                .documentPages(documentPageData)
-                .build();
-    }
-
-
     private File createTempFile(String filenamePrefix, String filenameSuffix) throws IOException {
 
         File tempFile = Files.createTempFile(filenamePrefix, filenameSuffix).toFile();
@@ -134,4 +114,30 @@ public class LayoutParsingStorageService {
         }
     }
 
+
+    public void storeSimplifiedText(LayoutParsingRequest layoutParsingRequest, SimplifiedText simplifiedText) {
+
+        storageService.storeJSONObject(TenantContext.getTenantId(), layoutParsingRequest.simplifiedTextStorageId(), simplifiedText);
+    }
+
+
+    @SneakyThrows
+    private InputStream getObject(String storageId) {
+
+        File tempFile = File.createTempFile("temp", ".data");
+        storageService.downloadTo(TenantContext.getTenantId(), storageId, tempFile);
+        Path path = Paths.get(tempFile.getPath());
+        return Files.newInputStream(path, StandardOpenOption.DELETE_ON_CLOSE);
+    }
+
+
+    @SneakyThrows
+    public void storeViewerDocument(LayoutParsingRequest layoutParsingRequest, ByteArrayOutputStream out) {
+
+        try (var in = new ByteArrayInputStream(out.toByteArray())) {
+
+            storageService.storeObject(TenantContext.getTenantId(), layoutParsingRequest.viewerDocumentStorageId(), in);
+        }
+    }
+
 }

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

@@ -72,9 +72,30 @@ public abstract class AbstractPageBlock {
     }
 
 
-    public boolean intersectsY(AbstractPageBlock atc) {
+    public boolean intersectsY(AbstractPageBlock apb) {
 
-        return this.minY <= atc.getMaxY() && this.maxY >= atc.getMinY();
+        return this.minY <= apb.getMaxY() && this.maxY >= apb.getMinY();
     }
 
+
+    public boolean almostIntersects(AbstractPageBlock apb, float yThreshold, float xThreshold) {
+
+        return this.almostIntersectsX(apb, xThreshold) && this.almostIntersectsY(apb, yThreshold);
+    }
+
+
+    private boolean almostIntersectsY(AbstractPageBlock apb, float threshold) {
+
+        return this.minY - threshold <= apb.getMaxY() && this.maxY + threshold >= apb.getMinY();
+    }
+
+
+    private boolean almostIntersectsX(AbstractPageBlock apb, float threshold) {
+
+        return this.minX - threshold <= apb.getMaxX() && this.maxX + threshold >= apb.getMinX();
+    }
+
+
+    public abstract boolean isEmpty();
+
 }

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

@@ -3,7 +3,6 @@ package com.knecon.fforesight.service.layoutparser.processor.model;
 import java.util.ArrayList;
 import java.util.List;
 
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.section.SectionGrid;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.StringFrequencyCounter;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.UnclassifiedText;
 
@@ -25,7 +24,6 @@ public class ClassificationDocument {
     private StringFrequencyCounter fontStyleCounter = new StringFrequencyCounter();
     private boolean headlines;
 
-    private SectionGrid sectionGrid = new SectionGrid();
     private long rulesVersion;
 
 }

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

@@ -1,15 +1,20 @@
 package com.knecon.fforesight.service.layoutparser.processor.model;
 
+import java.awt.geom.Rectangle2D;
 import java.util.ArrayList;
+import java.util.HashMap;
 import java.util.List;
+import java.util.Map;
 
 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
 import com.knecon.fforesight.service.layoutparser.processor.model.image.ClassifiedImage;
+import com.knecon.fforesight.service.layoutparser.processor.model.table.CleanRulings;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.StringFrequencyCounter;
 
 import lombok.Data;
 import lombok.NonNull;
 import lombok.RequiredArgsConstructor;
+import org.apache.pdfbox.pdmodel.documentinterchange.markedcontent.PDMarkedContent;
 
 @Data
 @RequiredArgsConstructor
@@ -35,4 +40,8 @@ public class ClassificationPage {
     private float pageWidth;
     private float pageHeight;
 
+    CleanRulings cleanRulings;
+
+    private Map<String, List<Rectangle2D>> markedContentBboxPerType = new HashMap<>();
+
 }

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

@@ -2,6 +2,7 @@ package com.knecon.fforesight.service.layoutparser.processor.model;
 
 import java.util.ArrayList;
 import java.util.List;
+import java.util.stream.Collectors;
 
 import com.knecon.fforesight.service.layoutparser.processor.model.image.ClassifiedImage;
 import com.knecon.fforesight.service.layoutparser.processor.model.table.TablePageBlock;
@@ -29,4 +30,10 @@ public class ClassificationSection {
         return tables;
     }
 
+
+    public List<AbstractPageBlock> getNonEmptyPageBlocks() {
+
+        return pageBlocks.stream().filter(pageBlock -> !pageBlock.isEmpty()).collect(Collectors.toList());
+    }
+
 }

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

@@ -3,6 +3,7 @@ package com.knecon.fforesight.service.layoutparser.processor.model;
 import java.awt.geom.Rectangle2D;
 import java.util.List;
 
+import com.knecon.fforesight.service.layoutparser.processor.model.table.Ruling;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
 
 import lombok.AllArgsConstructor;
@@ -17,5 +18,5 @@ public class PageContents {
     List<TextPositionSequence> sortedTextPositionSequences;
     Rectangle2D cropBox;
     Rectangle2D mediaBox;
-
+    List<Ruling> rulings;
 }

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

@@ -100,6 +100,11 @@ public class Document implements GenericSemanticNode {
     }
 
 
+    public Map<NodeType, Long> buildSemanticNodeCounts() {
+
+        return streamAllSubNodes().collect(Collectors.groupingBy(SemanticNode::getType, Collectors.counting()));
+    }
+
     @Override
     public String toString() {
 

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

@@ -34,7 +34,6 @@ public class Table implements SemanticNode {
 
     int numberOfRows;
     int numberOfCols;
-
     TextBlock textBlock;
 
     @Builder.Default
@@ -208,7 +207,6 @@ public class Table implements SemanticNode {
         return IntStream.range(0, numberOfCols).boxed().map(col -> getCell(row, col));
     }
 
-
     /**
      * Streams all TableCells row-wise and filters them with header == true.
      *

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

@@ -76,16 +76,22 @@ public class TableCell implements GenericSemanticNode {
         }
 
         if (textBlock == null) {
-            textBlock = streamAllSubNodes().filter(SemanticNode::isLeaf).map(SemanticNode::getLeafTextBlock).collect(new TextBlockCollector());
+            textBlock = buildTextBlock();
         }
         return textBlock;
     }
 
 
+    private TextBlock buildTextBlock() {
+
+        return streamAllSubNodes().filter(SemanticNode::isLeaf).map(SemanticNode::getLeafTextBlock).collect(new TextBlockCollector());
+    }
+
+
     @Override
     public String toString() {
 
-        return treeId + ": " + NodeType.TABLE_CELL + ": " + this.getTextBlock().buildSummary();
+        return treeId + ": " + NodeType.TABLE_CELL + ": " + this.buildTextBlock().buildSummary();
     }
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/table/Ruling.java

@@ -28,6 +28,7 @@ public class Ruling extends Line2D.Float {
         super(p1, p2);
     }
 
+
     public Ruling straightenVertical() {
 
         double y1 = Math.min(getY1(), getY2());
@@ -36,6 +37,7 @@ public class Ruling extends Line2D.Float {
         return new Ruling(new Point2D.Double(x, y1), new Point2D.Double(x, y2));
     }
 
+
     public Ruling straightenHorizontal() {
 
         double x1 = Math.min(getX1(), getX2());
@@ -444,6 +446,16 @@ public class Ruling extends Line2D.Float {
     }
 
 
+    public boolean almostMatches(Ruling ruling) {
+
+        final float TOLERANCE = 1;
+        return Math.abs(ruling.getX1() - x1) < TOLERANCE &&//
+                Math.abs(ruling.getY1() - y1) < TOLERANCE &&//
+                Math.abs(ruling.getX2() - x2) < TOLERANCE &&//
+                Math.abs(ruling.getY2() - y2) < TOLERANCE;
+    }
+
+
     private enum SOType {
         VERTICAL,
         HRIGHT,

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/table/TablePageBlock.java

@@ -1,12 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.model.table;
 
 import java.awt.geom.Point2D;
+import java.awt.geom.Rectangle2D;
 import java.util.ArrayList;
 import java.util.Collections;
 import java.util.HashSet;
 import java.util.List;
 import java.util.Set;
 import java.util.TreeMap;
+import java.util.stream.Collectors;
 
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
@@ -42,6 +44,12 @@ public class TablePageBlock extends AbstractPageBlock {
     }
 
 
+    @Override
+    public boolean isEmpty() {
+
+        return getColCount() == 0 || getRowCount() == 0;
+    }
+
     public List<List<Cell>> getRows() {
 
         if (rows == null) {
@@ -246,7 +254,8 @@ public class TablePageBlock extends AbstractPageBlock {
                 if (prevY != null && prevX != null) {
                     var cell = new Cell(new Point2D.Float(prevX, prevY), new Point2D.Float(x, y));
 
-                    var intersectionCell = cells.stream().filter(c -> cell.intersects(c) && cell.overlapRatio(c) > 0.1f).findFirst();
+                    var intersectionCell = cells.stream().filter(c -> intersects(cell, c)).findFirst();
+
                     intersectionCell.ifPresent(value -> cell.getTextBlocks().addAll(value.getTextBlocks()));
                     if (cell.hasMinimumSize()) {
                         row.add(cell);
@@ -267,6 +276,21 @@ public class TablePageBlock extends AbstractPageBlock {
     }
 
 
+
+    public boolean intersects(Cell cell1, Cell cell2) {
+        if (cell1.getHeight() <= 0 || cell2.getHeight() <= 0) {
+            return false;
+        }
+        double x0 = cell1.getX() + 2;
+        double y0 = cell1.getY() + 2;
+        return (cell2.x + cell2.width > x0 &&
+                cell2.y + cell2.height > y0 &&
+                cell2.x < x0 + cell1.getWidth() -2 &&
+                cell2.y < y0 + cell1.getHeight() -2);
+    }
+
+
+
     @Override
     public String getText() {
 
@@ -304,6 +328,8 @@ public class TablePageBlock extends AbstractPageBlock {
     }
 
 
+
+
     public String getTextAsHtml() {
 
         StringBuilder sb = new StringBuilder();

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

@@ -17,7 +17,6 @@ import lombok.SneakyThrows;
 @AllArgsConstructor
 public class RedTextPosition {
 
-    private String textMatrix;
     private float[] position;
 
     @JsonIgnore
@@ -56,8 +55,6 @@ public class RedTextPosition {
 
         pos.setFontSizeInPt(textPosition.getFontSizeInPt());
 
-        pos.setTextMatrix(textPosition.getTextMatrix().toString());
-
         var position = new float[4];
 
         position[0] = textPosition.getXDirAdj();

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

@@ -1,17 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.model.text;
-
-import lombok.AllArgsConstructor;
-import lombok.Builder;
-import lombok.Data;
-import lombok.NoArgsConstructor;
-
-@Data
-@Builder
-@NoArgsConstructor
-@AllArgsConstructor
-public class SimplifiedSectionText {
-
-    private int sectionNumber;
-    private String text;
-
-}

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

@@ -1,20 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.model.text;
-
-import java.util.ArrayList;
-import java.util.List;
-
-import lombok.AllArgsConstructor;
-import lombok.Builder;
-import lombok.Data;
-import lombok.NoArgsConstructor;
-
-@Data
-@Builder
-@NoArgsConstructor
-@AllArgsConstructor
-public class SimplifiedText {
-
-    private int numberOfPages;
-    private List<SimplifiedSectionText> sectionTexts = new ArrayList<>();
-
-}

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

@@ -73,6 +73,7 @@ public class TextPageBlock extends AbstractPageBlock {
 
         return sequences.get(0).getPageWidth();
     }
+    
 
     public static TextPageBlock merge(List<TextPageBlock> textBlocksToMerge) {
 
@@ -364,4 +365,11 @@ public class TextPageBlock extends AbstractPageBlock {
 
     }
 
+
+    @Override
+    public boolean isEmpty() {
+
+        return sequences.isEmpty();
+    }
+
 }

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

@@ -196,6 +196,12 @@ public class TextPositionSequence implements CharSequence {
     }
 
 
+    public float getTextHeightNoPadding() {
+
+        return textPositions.get(0).getHeightDir();
+    }
+
+
     @JsonIgnore
     @JsonAttribute(ignore = true)
     public float getTextHeight() {
@@ -224,6 +230,9 @@ public class TextPositionSequence implements CharSequence {
     @JsonAttribute(ignore = true)
     public String getFont() {
 
+        if (textPositions.get(0).getFontName() == null) {
+            return "none";
+        }
         return textPositions.get(0).getFontName().toLowerCase(Locale.ROOT).replaceAll(",bold", "").replaceAll(",italic", "");
     }
 
@@ -232,6 +241,9 @@ public class TextPositionSequence implements CharSequence {
     @JsonAttribute(ignore = true)
     public String getFontStyle() {
 
+        if (textPositions.get(0).getFontName() == null) {
+            return "standard";
+        }
         String lowercaseFontName = textPositions.get(0).getFontName().toLowerCase(Locale.ROOT);
 
         if (lowercaseFontName.contains("bold") && lowercaseFontName.contains("italic")) {
@@ -243,7 +255,6 @@ public class TextPositionSequence implements CharSequence {
         } else {
             return "standard";
         }
-
     }
 
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/visualization/ColoredLine.java

@@ -0,0 +1,8 @@
+package com.knecon.fforesight.service.layoutparser.processor.model.visualization;
+
+import java.awt.Color;
+import java.awt.geom.Line2D;
+
+public record ColoredLine(Line2D line, Color color) {
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/visualization/ColoredRectangle.java

@@ -0,0 +1,8 @@
+package com.knecon.fforesight.service.layoutparser.processor.model.visualization;
+
+import java.awt.Color;
+import java.awt.geom.Rectangle2D;
+
+public record ColoredRectangle(Rectangle2D rectangle2D, Color color) {
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/visualization/FilledRectangle.java

@@ -0,0 +1,8 @@
+package com.knecon.fforesight.service.layoutparser.processor.model.visualization;
+
+import java.awt.Color;
+import java.awt.geom.Rectangle2D;
+
+public record FilledRectangle(Rectangle2D rectangle2D, Color color, float alpha) {
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/visualization/LayoutGrid.java

@@ -0,0 +1,27 @@
+package com.knecon.fforesight.service.layoutparser.processor.model.visualization;
+
+import java.util.ArrayList;
+import java.util.List;
+
+import lombok.AccessLevel;
+import lombok.Getter;
+import lombok.experimental.FieldDefaults;
+
+@Getter
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class LayoutGrid {
+
+    int numberOfPages;
+    List<VisualizationsOnPage> visualizationsPerPages;
+
+
+    public LayoutGrid(int numberOfPages) {
+
+        this.numberOfPages = numberOfPages;
+        this.visualizationsPerPages = new ArrayList<>(numberOfPages);
+        for (int i = 0; i < numberOfPages; i++) {
+            this.visualizationsPerPages.add(VisualizationsOnPage.builder().pageNumber(i).build());
+        }
+    }
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/visualization/PlacedText.java

@@ -0,0 +1,7 @@
+package com.knecon.fforesight.service.layoutparser.processor.model.visualization;
+
+import java.awt.geom.Point2D;
+
+public record PlacedText(String text, Point2D lineStart) {
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/model/visualization/VisualizationsOnPage.java

@@ -0,0 +1,26 @@
+package com.knecon.fforesight.service.layoutparser.processor.model.visualization;
+
+import java.util.LinkedList;
+import java.util.List;
+
+import lombok.AccessLevel;
+import lombok.Builder;
+import lombok.Getter;
+import lombok.experimental.FieldDefaults;
+
+@Getter
+@Builder
+@FieldDefaults(makeFinal = true, level = AccessLevel.PRIVATE)
+public class VisualizationsOnPage {
+
+    int pageNumber;
+    @Builder.Default
+    List<PlacedText> placedTexts = new LinkedList<>();
+    @Builder.Default
+    List<ColoredLine> coloredLines = new LinkedList<>();
+    @Builder.Default
+    List<ColoredRectangle> coloredRectangles = new LinkedList<>();
+    @Builder.Default
+    List<FilledRectangle> filledRectangles = new LinkedList<>();
+
+}

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/adapter/CvTableParsingAdapter.java

@@ -8,6 +8,7 @@ import java.util.Map;
 
 import org.springframework.stereotype.Service;
 
+import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.PageInfo;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableCells;
 import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
 
@@ -24,24 +25,26 @@ public class CvTableParsingAdapter {
         Map<Integer, List<TableCells>> tableCells = new HashMap<>();
         tableServiceResponse.getData()
                 .forEach(tableData -> tableCells.computeIfAbsent(tableData.getPageInfo().getNumber(), tableCell -> new ArrayList<>())
-                        .addAll(convertTableCells(tableData.getTableCells())));
+                        .addAll(convertTableCells(tableData.getTableCells(), tableData.getPageInfo())));
 
         return tableCells;
     }
 
 
-    private Collection<TableCells> convertTableCells(List<TableCells> tableCells) {
+    private Collection<TableCells> convertTableCells(List<TableCells> tableCells, PageInfo pageInfo) {
 
         List<TableCells> cvParsedTableCells = new ArrayList<>();
 
-        tableCells.forEach(t -> cvParsedTableCells.add(TableCells.builder()
-                .y0(t.getY0())
-                .x1(t.getX1())
-                .y1(t.getY1())
-                .x0(t.getX0())
-                .width(t.getWidth())
-                .height(t.getHeight())
-                .build()));
+        tableCells.stream()
+                .filter(cell -> cell.getWidth() < pageInfo.getWidth() * 0.98 && cell.getHeight() < pageInfo.getHeight() * 0.98)
+                .forEach(t -> cvParsedTableCells.add(TableCells.builder()
+                        .y0(t.getY0())
+                        .x1(t.getX1())
+                        .y1(t.getY1())
+                        .x0(t.getX0())
+                        .width(t.getWidth())
+                        .height(t.getHeight())
+                        .build()));
 
         return cvParsedTableCells;
     }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/image/Classification.java

@@ -3,9 +3,15 @@ package com.knecon.fforesight.service.layoutparser.processor.python_api.model.im
 import java.util.HashMap;
 import java.util.Map;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class Classification {
 
     private Map<String, Float> probabilities = new HashMap<>();

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/image/FilterGeometry.java

@@ -1,8 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.python_api.model.image;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class FilterGeometry {
 
     private ImageSize imageSize;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/image/Filters.java

@@ -1,8 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.python_api.model.image;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class Filters {
 
     private FilterGeometry geometry;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/image/Geometry.java

@@ -1,8 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.python_api.model.image;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class Geometry {
 
     private float width;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/image/ImageFormat.java

@@ -1,8 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.python_api.model.image;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class ImageFormat {
 
     private float quotient;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/image/ImageMetadata.java

@@ -1,8 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.python_api.model.image;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class ImageMetadata {
 
     private Classification classification;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/image/ImageServiceResponse.java

@@ -6,9 +6,15 @@ import java.util.List;
 import com.fasterxml.jackson.annotation.JsonAlias;
 import com.fasterxml.jackson.annotation.JsonProperty;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class ImageServiceResponse {
 
     private String dossierId;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/image/ImageSize.java

@@ -1,8 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.python_api.model.image;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class ImageSize {
 
     private float quotient;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/image/Position.java

@@ -1,8 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.python_api.model.image;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class Position {
 
     private float x1;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/image/Probability.java

@@ -1,8 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.python_api.model.image;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class Probability {
 
     private boolean unconfident;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/table/PageInfo.java

@@ -1,8 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.python_api.model.table;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class PageInfo {
 
     private int number;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/table/PdfTableCell.java

@@ -3,12 +3,12 @@ package com.knecon.fforesight.service.layoutparser.processor.python_api.model.ta
 import lombok.AllArgsConstructor;
 import lombok.Builder;
 import lombok.Data;
-import lombok.RequiredArgsConstructor;
+import lombok.NoArgsConstructor;
 
 @Data
 @Builder
+@NoArgsConstructor
 @AllArgsConstructor
-@RequiredArgsConstructor
 public class PdfTableCell {
 
     private float x0;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/table/TableCells.java

@@ -1,10 +1,14 @@
 package com.knecon.fforesight.service.layoutparser.processor.python_api.model.table;
 
+import lombok.AllArgsConstructor;
 import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
 @Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class TableCells {
 
     private float x0;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/table/TableData.java

@@ -3,9 +3,15 @@ package com.knecon.fforesight.service.layoutparser.processor.python_api.model.ta
 import java.util.ArrayList;
 import java.util.List;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class TableData {
 
     private PageInfo pageInfo;

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/python_api/model/table/TableServiceResponse.java

@@ -3,9 +3,15 @@ package com.knecon.fforesight.service.layoutparser.processor.python_api.model.ta
 import java.util.ArrayList;
 import java.util.List;
 
+import lombok.AllArgsConstructor;
+import lombok.Builder;
 import lombok.Data;
+import lombok.NoArgsConstructor;
 
 @Data
+@Builder
+@NoArgsConstructor
+@AllArgsConstructor
 public class TableServiceResponse {
 
     private String dossierId;

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

@@ -1,40 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.queue;
-
-import static com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingQueueNames.LAYOUT_PARSING_FINISHED_EVENT_QUEUE;
-import static com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingQueueNames.LAYOUT_PARSING_DLQ;
-import static com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingQueueNames.LAYOUT_PARSING_REQUEST_QUEUE;
-
-import org.springframework.amqp.core.Queue;
-import org.springframework.amqp.core.QueueBuilder;
-import org.springframework.context.annotation.Bean;
-import org.springframework.context.annotation.Configuration;
-
-import lombok.RequiredArgsConstructor;
-
-@Configuration
-@RequiredArgsConstructor
-public class MessagingConfiguration {
-
-    @Bean
-    public Queue layoutparsingRequestQueue() {
-
-        return QueueBuilder.durable(LAYOUT_PARSING_REQUEST_QUEUE)//
-                .withArgument("x-dead-letter-exchange", "").withArgument("x-dead-letter-routing-key", LAYOUT_PARSING_DLQ).build();
-    }
-
-
-    @Bean
-    public Queue layoutparsingResponseQueue() {
-
-        return QueueBuilder.durable(LAYOUT_PARSING_FINISHED_EVENT_QUEUE)//
-                .withArgument("x-dead-letter-exchange", "").withArgument("x-dead-letter-routing-key", LAYOUT_PARSING_DLQ).build();
-    }
-
-
-    @Bean
-    public Queue layoutparsingDLQ() {
-
-        return QueueBuilder.durable(LAYOUT_PARSING_DLQ).build();
-    }
-
-}

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

@@ -1,5 +1,6 @@
 package com.knecon.fforesight.service.layoutparser.processor.services;
 
+import java.util.Comparator;
 import java.util.List;
 
 import org.springframework.stereotype.Service;
@@ -8,17 +9,74 @@ import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlo
 import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
 import com.knecon.fforesight.service.layoutparser.processor.model.FloatFrequencyCounter;
 import com.knecon.fforesight.service.layoutparser.processor.model.table.Cell;
+import com.knecon.fforesight.service.layoutparser.processor.model.table.Ruling;
 import com.knecon.fforesight.service.layoutparser.processor.model.table.TablePageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 import com.knecon.fforesight.service.layoutparser.processor.utils.PositionUtils;
 
 @Service
 public class BodyTextFrameService {
 
+    private static final float RULING_HEIGHT_THRESHOLD = 0.15f; // multiplied with page height. Header/Footer Rulings must be within that border of the page.
+    private static final float RULING_WIDTH_THRESHOLD = 0.75f; // multiplied with page width. Header/Footer Rulings must be at least that wide.
 
+    public void setBodyTextFrames(ClassificationDocument classificationDocument, LayoutParsingType layoutParsingType) {
+
+        Rectangle bodyTextFrame = calculateBodyTextFrame(classificationDocument.getPages(), classificationDocument.getFontSizeCounter(), false, layoutParsingType);
+        Rectangle landscapeBodyTextFrame = calculateBodyTextFrame(classificationDocument.getPages(), classificationDocument.getFontSizeCounter(), true, layoutParsingType);
+        for (ClassificationPage page : classificationDocument.getPages()) {
+//            var updatedBodyTextFrame = getBodyTextFrameFromRulings(page, bodyTextFrame, landscapeBodyTextFrame);
+            setBodyTextFrameAdjustedToPage(page, bodyTextFrame, landscapeBodyTextFrame);
+        }
+    }
+
+
+    private Rectangle getBodyTextFrameFromRulings(ClassificationPage page, Rectangle bodyTextFrame, Rectangle landscapeBodyTextFrame) {
+
+        List<Ruling> potentialFooterRulings = getPotentialFooterRulings(page);
+        List<Ruling> potentialHeaderRulings = getPotentialHeaderRulings(page);
+        var x = bodyTextFrame.getTopLeft().getX();
+        var y = bodyTextFrame.getTopLeft().getY();
+        var w = bodyTextFrame.getWidth();
+        var h = bodyTextFrame.getHeight();
+        if (!potentialFooterRulings.isEmpty()) {
+            h = y + h - potentialFooterRulings.get(0).getTop();
+            y = potentialFooterRulings.get(0).getTop();
+        }
+        if (!potentialHeaderRulings.isEmpty()) {
+            h = potentialHeaderRulings.get(0).getBottom() - bodyTextFrame.getTopLeft().getY();
+        }
+        return new Rectangle(new Point(x, y), w, h, page.getPageNumber());
+    }
+
+
+    private List<Ruling> getPotentialFooterRulings(ClassificationPage page) {
+
+        return page.getCleanRulings()
+                .getHorizontal()
+                .stream()
+                .filter(ruling -> ruling.getY1() < page.getPageHeight() * RULING_HEIGHT_THRESHOLD)
+                .filter(ruling -> ruling.getWidth() > RULING_WIDTH_THRESHOLD * page.getPageWidth())
+                .sorted(Comparator.comparingDouble(Ruling::getTop))
+                .toList();
+    }
+
+
+    private List<Ruling> getPotentialHeaderRulings(ClassificationPage page) {
+
+        return page.getCleanRulings()
+                .getHorizontal()
+                .stream()
+                .filter(ruling -> ruling.getY1() > page.getPageHeight() * (1 - RULING_HEIGHT_THRESHOLD))
+                .filter(ruling -> ruling.getWidth() > RULING_WIDTH_THRESHOLD * page.getPageWidth())
+                .sorted(Comparator.comparingDouble(Ruling::getBottom).reversed())
+                .toList();
+    }
 
 
     /**
@@ -34,7 +92,7 @@ public class BodyTextFrameService {
      * @param bodyTextFrame          frame that contains the main text on portrait pages
      * @param landscapeBodyTextFrame frame that contains the main text on landscape pages
      */
-    public void setBodyTextFrameAdjustedToPage(ClassificationPage page, Rectangle bodyTextFrame, Rectangle landscapeBodyTextFrame) {
+    private void setBodyTextFrameAdjustedToPage(ClassificationPage page, Rectangle bodyTextFrame, Rectangle landscapeBodyTextFrame) {
 
         Rectangle textFrame = page.isLandscape() ? landscapeBodyTextFrame : bodyTextFrame;
 
@@ -69,7 +127,10 @@ public class BodyTextFrameService {
      * @param landscape               Calculate for landscape or portrait
      * @return Rectangle of the text frame
      */
-    public Rectangle calculateBodyTextFrame(List<ClassificationPage> pages, FloatFrequencyCounter documentFontSizeCounter, boolean landscape, LayoutParsingType layoutParsingType) {
+    protected Rectangle calculateBodyTextFrame(List<ClassificationPage> pages,
+                                               FloatFrequencyCounter documentFontSizeCounter,
+                                               boolean landscape,
+                                               LayoutParsingType layoutParsingType) {
 
         float approximateHeaderLineCount;
         if (layoutParsingType.equals(LayoutParsingType.TAAS)) {
@@ -94,9 +155,14 @@ public class BodyTextFrameService {
                         continue;
                     }
 
+                    if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.HEADER)
+                            || MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.FOOTER)) {
+                        continue;
+                    }
+
                     float approxLineCount = PositionUtils.getApproxLineCount(textBlock);
-                    if (layoutParsingType.equals(LayoutParsingType.DOCUMINE) && approxLineCount < approximateHeaderLineCount && textBlock.getMaxY() >= page.getPageHeight() - (page.getPageHeight() / 10)
-                            || !layoutParsingType.equals(LayoutParsingType.DOCUMINE) && approxLineCount < approximateHeaderLineCount){
+                    if (layoutParsingType.equals(LayoutParsingType.DOCUMINE) && approxLineCount < approximateHeaderLineCount && textBlock.getMaxY() >= page.getPageHeight() - (page.getPageHeight() / 10) || !layoutParsingType.equals(
+                            LayoutParsingType.DOCUMINE) && approxLineCount < approximateHeaderLineCount) {
                         continue;
                     }
 
@@ -170,4 +236,4 @@ public class BodyTextFrameService {
 
     }
 
-}
+}

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

@@ -0,0 +1,73 @@
+package com.knecon.fforesight.service.layoutparser.processor.services;
+
+import java.io.IOException;
+import java.util.Collection;
+import java.util.LinkedList;
+import java.util.List;
+import java.util.Map;
+import java.util.stream.Collectors;
+
+import org.apache.pdfbox.Loader;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.springframework.core.io.ClassPathResource;
+
+import com.knecon.fforesight.service.layoutparser.processor.model.PageContents;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
+import com.knecon.fforesight.service.layoutparser.processor.services.parsing.PDFLinesTextStripper;
+import com.knecon.fforesight.service.layoutparser.processor.utils.RectangleTransformations;
+
+import lombok.experimental.UtilityClass;
+
+@UtilityClass
+public class PageContentExtractor {
+
+    public List<PageContents> getSortedPageContents(String filename) throws IOException {
+
+        List<PageContents> textPositionSequencesPerPage = new LinkedList<>();
+        ClassPathResource pdfResource = new ClassPathResource(filename);
+
+        try (PDDocument pdDocument = Loader.loadPDF(pdfResource.getFile())) {
+
+            for (int pageNumber = 1; pageNumber < pdDocument.getNumberOfPages() + 1; pageNumber++) {
+
+                PDFLinesTextStripper stripper = new PDFLinesTextStripper();
+                PDPage pdPage = pdDocument.getPage(pageNumber - 1);
+                stripper.setPageNumber(pageNumber);
+                stripper.setSortByPosition(true);
+                stripper.setStartPage(pageNumber);
+                stripper.setEndPage(pageNumber);
+                stripper.setPdpage(pdPage);
+                stripper.getText(pdDocument);
+
+                Map<Float, List<TextPositionSequence>> sortedTextPositionSequencesPerDir = stripper.getTextPositionSequences()
+                        .stream()
+                        .collect(Collectors.groupingBy(textPositionSequence -> textPositionSequence.getDir().getDegrees()));
+
+                var sortedTextPositionSequences = sortByDirAccordingToPageRotation(sortedTextPositionSequencesPerDir, pdPage.getRotation());
+
+                textPositionSequencesPerPage.add(new PageContents(sortedTextPositionSequences,
+                        RectangleTransformations.toRectangle2D(pdPage.getCropBox()),
+                        RectangleTransformations.toRectangle2D(pdPage.getMediaBox()),
+                        stripper.getRulings()));
+            }
+        }
+    
+        return textPositionSequencesPerPage;
+    }
+
+
+    public List<TextPositionSequence> sortByDirAccordingToPageRotation(Map<Float, List<TextPositionSequence>> sortedTextPositionSequencesPerDir, int rotation) {
+
+        LinkedList<Float> sortedKeys = new LinkedList<>(sortedTextPositionSequencesPerDir.keySet().stream().sorted().toList());
+
+        for (int i = 0; i < sortedKeys.size(); i++) {
+            if (sortedKeys.get(i) < rotation) {
+                Float keyToSwap = sortedKeys.remove(i);
+                sortedKeys.addLast(keyToSwap);
+            }
+        }
+        return sortedKeys.stream().map(sortedTextPositionSequencesPerDir::get).flatMap(Collection::stream).toList();
+    }
+
+}

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

@@ -1,151 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.services;
-
-import java.util.ArrayList;
-import java.util.List;
-import java.util.Map;
-
-import org.apache.pdfbox.pdmodel.PDDocument;
-import org.apache.pdfbox.pdmodel.PDPage;
-import org.apache.pdfbox.pdmodel.common.PDRectangle;
-import org.springframework.stereotype.Service;
-
-import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
-import com.knecon.fforesight.service.layoutparser.processor.python_api.adapter.ImageServiceResponseAdapter;
-import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableCells;
-import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
-import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
-import com.knecon.fforesight.service.layoutparser.processor.model.image.ClassifiedImage;
-import com.knecon.fforesight.service.layoutparser.processor.model.table.CleanRulings;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
-import com.knecon.fforesight.service.layoutparser.processor.services.parsing.PDFLinesTextStripper;
-import com.knecon.fforesight.service.layoutparser.processor.services.blockification.DocuMineBlockificationService;
-import com.knecon.fforesight.service.layoutparser.processor.services.blockification.RedactManagerBlockificationService;
-import com.knecon.fforesight.service.layoutparser.processor.services.blockification.TaasBlockificationService;
-
-import lombok.RequiredArgsConstructor;
-import lombok.SneakyThrows;
-import lombok.extern.slf4j.Slf4j;
-
-@Slf4j
-@Service
-@RequiredArgsConstructor
-public class PdfParsingService {
-
-    private final RulingCleaningService rulingCleaningService;
-    private final TableExtractionService tableExtractionService;
-    private final ImageServiceResponseAdapter imageServiceResponseAdapter;
-    private final TaasBlockificationService taasBlockificationService;
-    private final DocuMineBlockificationService docuMineBlockificationService;
-    private final RedactManagerBlockificationService redactManagerBlockificationService;
-
-
-    public ClassificationDocument parseDocument(LayoutParsingType layoutParsingType,
-                                                PDDocument originDocument,
-                                                Map<Integer, List<TableCells>> pdfTableCells,
-                                                Map<Integer, List<ClassifiedImage>> pdfImages) {
-
-        ClassificationDocument document = new ClassificationDocument();
-        List<ClassificationPage> classificationPages = new ArrayList<>();
-
-        originDocument.setAllSecurityToBeRemoved(true);
-        long pageCount = originDocument.getNumberOfPages();
-
-        for (int pageNumber = 1; pageNumber <= pageCount; pageNumber++) {
-            parsePage(layoutParsingType, pdfImages, originDocument, pdfTableCells, document, classificationPages, pageNumber);
-        }
-
-        document.setPages(classificationPages);
-
-        return document;
-    }
-
-
-    @SneakyThrows
-    private void parsePage(LayoutParsingType layoutParsingType,
-                           Map<Integer, List<ClassifiedImage>> pdfImages,
-                           PDDocument pdDocument,
-                           Map<Integer, List<TableCells>> pdfTableCells,
-                           ClassificationDocument document,
-                           List<ClassificationPage> classificationPages,
-                           int pageNumber) {
-
-        PDFLinesTextStripper stripper = new PDFLinesTextStripper();
-        PDPage pdPage = pdDocument.getPage(pageNumber - 1);
-        stripper.setPageNumber(pageNumber);
-        stripper.setStartPage(pageNumber);
-        stripper.setEndPage(pageNumber);
-        stripper.setPdpage(pdPage);
-        stripper.getText(pdDocument);
-
-        PDRectangle pdr = pdPage.getMediaBox();
-
-        int rotation = pdPage.getRotation();
-        boolean isLandscape = pdr.getWidth() > pdr.getHeight() && (rotation == 0 || rotation == 180) || pdr.getHeight() > pdr.getWidth() && (rotation == 90 || rotation == 270);
-
-        PDRectangle cropbox = pdPage.getCropBox();
-        CleanRulings cleanRulings = rulingCleaningService.getCleanRulings(pdfTableCells.get(pageNumber),
-                stripper.getRulings(),
-                stripper.getMinCharWidth(),
-                stripper.getMaxCharHeight());
-
-        ClassificationPage classificationPage = switch (layoutParsingType) {
-            case REDACT_MANAGER -> redactManagerBlockificationService.blockify(stripper.getTextPositionSequences(), cleanRulings.getHorizontal(), cleanRulings.getVertical());
-            case TAAS -> taasBlockificationService.blockify(stripper.getTextPositionSequences(), cleanRulings.getHorizontal(), cleanRulings.getVertical());
-            case DOCUMINE -> docuMineBlockificationService.blockify(stripper.getTextPositionSequences(), cleanRulings.getHorizontal(), cleanRulings.getVertical());
-        };
-
-        classificationPage.setRotation(rotation);
-        classificationPage.setLandscape(isLandscape);
-        classificationPage.setPageNumber(pageNumber);
-        classificationPage.setPageWidth(cropbox.getWidth());
-        classificationPage.setPageHeight(cropbox.getHeight());
-
-        // If images is ocr needs to be calculated before textBlocks are moved into tables, otherwise findOcr algorithm needs to be adopted.
-        if (pdfImages != null && pdfImages.containsKey(pageNumber)) {
-            classificationPage.setImages(pdfImages.get(pageNumber));
-            imageServiceResponseAdapter.findOcr(classificationPage);
-        }
-
-        tableExtractionService.extractTables(cleanRulings, classificationPage, layoutParsingType);
-        buildPageStatistics(classificationPage);
-        increaseDocumentStatistics(classificationPage, document);
-
-        classificationPages.add(classificationPage);
-    }
-
-
-    private void increaseDocumentStatistics(ClassificationPage classificationPage, ClassificationDocument document) {
-
-        if (!classificationPage.isLandscape()) {
-            document.getFontSizeCounter().addAll(classificationPage.getFontSizeCounter().getCountPerValue());
-        }
-        document.getFontCounter().addAll(classificationPage.getFontCounter().getCountPerValue());
-        document.getTextHeightCounter().addAll(classificationPage.getTextHeightCounter().getCountPerValue());
-        document.getFontStyleCounter().addAll(classificationPage.getFontStyleCounter().getCountPerValue());
-    }
-
-
-    private void buildPageStatistics(ClassificationPage classificationPage) {
-
-        // Collect all statistics for the classificationPage, except from blocks inside tables, as tables will always be added to BodyTextFrame.
-        for (AbstractPageBlock textBlock : classificationPage.getTextBlocks()) {
-            if (textBlock instanceof TextPageBlock) {
-                if (((TextPageBlock) textBlock).getSequences() == null) {
-                    continue;
-                }
-                for (TextPositionSequence word : ((TextPageBlock) textBlock).getSequences()) {
-                    classificationPage.getTextHeightCounter().add(word.getTextHeight());
-                    classificationPage.getFontCounter().add(word.getFont());
-                    classificationPage.getFontSizeCounter().add(word.getFontSize());
-                    classificationPage.getFontStyleCounter().add(word.getFontStyle());
-                }
-            }
-        }
-
-    }
-
-}
-
-

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

@@ -12,9 +12,9 @@ import java.util.Map;
 
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableCells;
 import com.knecon.fforesight.service.layoutparser.processor.model.table.CleanRulings;
 import com.knecon.fforesight.service.layoutparser.processor.model.table.Ruling;
+import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableCells;
 import com.knecon.fforesight.service.layoutparser.processor.utils.DoubleComparisons;
 
 import lombok.RequiredArgsConstructor;
@@ -25,10 +25,13 @@ import lombok.extern.slf4j.Slf4j;
 @RequiredArgsConstructor
 public class RulingCleaningService {
 
-    public CleanRulings getCleanRulings(List<TableCells> tableCells, List<Ruling> rulings, float minCharWidth, float maxCharHeight) {
+    private static final float THRESHOLD = 6;
+
+
+    public CleanRulings getCleanRulings(List<TableCells> tableCells, List<Ruling> rulings) {
 
         if (!rulings.isEmpty()) {
-            snapPoints(rulings, minCharWidth, maxCharHeight);
+            snapPoints(rulings);
         }
 
         List<Ruling> vrs = new ArrayList<>();
@@ -53,14 +56,11 @@ public class RulingCleaningService {
         }
         List<Ruling> horizontalRulingLines = collapseOrientedRulings(hrs);
 
-        return CleanRulings.builder()
-                .vertical(verticalRulingLines)
-                .horizontal(horizontalRulingLines)
-                .build();
+        return CleanRulings.builder().vertical(verticalRulingLines).horizontal(horizontalRulingLines).build();
     }
 
 
-    public void snapPoints(List<? extends Line2D.Float> rulings, float xThreshold, float yThreshold) {
+    public void snapPoints(List<? extends Line2D.Float> rulings) {
 
         // collect points and keep a Line -> p1,p2 map
         Map<Line2D.Float, Point2D[]> linesToPoints = new HashMap<>();
@@ -81,7 +81,7 @@ public class RulingCleaningService {
 
         for (Point2D p : points.subList(1, points.size() - 1)) {
             List<Point2D> last = groupedPoints.get(groupedPoints.size() - 1);
-            if (Math.abs(p.getX() - last.get(0).getX()) < xThreshold) {
+            if (Math.abs(p.getX() - last.get(0).getX()) < THRESHOLD) {
                 groupedPoints.get(groupedPoints.size() - 1).add(p);
             } else {
                 groupedPoints.add(new ArrayList<>(Collections.singletonList(p)));
@@ -108,7 +108,7 @@ public class RulingCleaningService {
 
         for (Point2D p : points.subList(1, points.size() - 1)) {
             List<Point2D> last = groupedPoints.get(groupedPoints.size() - 1);
-            if (Math.abs(p.getY() - last.get(0).getY()) < yThreshold) {
+            if (Math.abs(p.getY() - last.get(0).getY()) < THRESHOLD) {
                 groupedPoints.get(groupedPoints.size() - 1).add(p);
             } else {
                 groupedPoints.add(new ArrayList<>(Collections.singletonList(p)));

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

@@ -1,146 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.services;
-
-import java.awt.geom.Rectangle2D;
-import java.util.HashMap;
-import java.util.List;
-import java.util.Map;
-import java.util.Objects;
-import java.util.Set;
-import java.util.function.BiConsumer;
-import java.util.function.BinaryOperator;
-import java.util.function.Function;
-import java.util.function.Supplier;
-import java.util.stream.Collector;
-import java.util.stream.Stream;
-
-import org.springframework.stereotype.Service;
-
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Point;
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.section.CellRectangle;
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.section.SectionGrid;
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.section.SectionRectangle;
-import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Page;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SemanticNode;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Table;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.TableCell;
-
-import lombok.RequiredArgsConstructor;
-
-@Service
-@RequiredArgsConstructor
-public class SectionGridCreatorService {
-
-    public SectionGrid createSectionGrid(Document document) {
-
-        Map<Integer, List<SectionRectangle>> sectionBBox = document.streamAllSubNodesOfType(NodeType.SECTION).map(SemanticNode::getBBox).collect(new SectionGridCollector());
-        Map<Integer, List<SectionRectangle>> paragraphBBox = document.streamAllSubNodesOfType(NodeType.PARAGRAPH).map(SemanticNode::getBBox).collect(new SectionGridCollector());
-        Map<Integer, List<SectionRectangle>> headlineBBox = document.streamAllSubNodesOfType(NodeType.HEADLINE).map(SemanticNode::getBBox).collect(new SectionGridCollector());
-        Map<Integer, List<SectionRectangle>> tableBBox = document.streamAllSubNodesOfType(NodeType.TABLE).map(node -> (Table) node).collect(new TableGridCollector());
-        var sectionGrid = new SectionGrid();
-
-        sectionGrid.setRectanglesPerPage(mergeMapsByConcatenatingLists(//
-                mergeMapsByConcatenatingLists(paragraphBBox, headlineBBox), //
-                mergeMapsByConcatenatingLists(sectionBBox, tableBBox)));
-
-        return sectionGrid;
-    }
-
-
-    private static abstract class GridCollector<T> implements Collector<T, Map<Integer, List<SectionRectangle>>, Map<Integer, List<SectionRectangle>>> {
-
-        @Override
-        public Supplier<Map<Integer, List<SectionRectangle>>> supplier() {
-
-            return HashMap::new;
-        }
-
-
-        @Override
-        public Function<Map<Integer, List<SectionRectangle>>, Map<Integer, List<SectionRectangle>>> finisher() {
-
-            return Function.identity();
-        }
-
-
-        @Override
-        public BinaryOperator<Map<Integer, List<SectionRectangle>>> combiner() {
-
-            return SectionGridCreatorService::mergeMapsByConcatenatingLists;
-        }
-
-
-        @Override
-        public Set<Characteristics> characteristics() {
-
-            return Set.of(Characteristics.IDENTITY_FINISH, Characteristics.CONCURRENT, Characteristics.UNORDERED);
-        }
-
-    }
-
-    private static class TableGridCollector extends GridCollector<Table> {
-
-        @Override
-        public BiConsumer<Map<Integer, List<SectionRectangle>>, Table> accumulator() {
-
-            return (map, table) -> table.getPages()
-                    .forEach(page -> map.merge(page.getNumber(), List.of(toSectionRectangle(table, page, table.getPages().size())), SectionGridCreatorService::concatLists));
-        }
-
-
-        private static SectionRectangle toSectionRectangle(Table table, Page page, int numberOfParts) {
-
-            Rectangle2D rect = table.getBBox().get(page);
-            List<CellRectangle> tableCellRectangles = table.streamTableCells()
-                    .map(TableCell::getBBox)
-                    .map(map -> map.get(page))
-                    .filter(Objects::nonNull)
-                    .map(rectangle2D -> new CellRectangle(new Point((float) rectangle2D.getX(), (float) rectangle2D.getY()),
-                            (float) rectangle2D.getWidth(),
-                            (float) rectangle2D.getHeight()))
-                    .toList();
-            return new SectionRectangle(new Point((float) rect.getX(), (float) rect.getY()),
-                    (float) rect.getWidth(),
-                    (float) rect.getHeight(),
-                    1,
-                    numberOfParts,
-                    tableCellRectangles);
-        }
-
-    }
-
-    private static class SectionGridCollector extends GridCollector<Map<Page, Rectangle2D>> {
-
-        @Override
-        public BiConsumer<Map<Integer, List<SectionRectangle>>, Map<Page, Rectangle2D>> accumulator() {
-
-            return (mapToKeep, mapToMerge) -> mapToMerge.forEach((page, rectangle) -> mapToKeep.merge(page.getNumber(),
-                    List.of(toSectionRectangle(rectangle, mapToMerge.values().size())),
-                    SectionGridCreatorService::concatLists));
-
-        }
-
-
-        private static SectionRectangle toSectionRectangle(Rectangle2D rect, int numberOfParts) {
-
-            return new SectionRectangle(new Point((float) rect.getX(), (float) rect.getY()), (float) rect.getWidth(), (float) rect.getHeight(), 1, numberOfParts, null);
-        }
-
-    }
-
-
-    private static Map<Integer, List<SectionRectangle>> mergeMapsByConcatenatingLists(Map<Integer, List<SectionRectangle>> mapToKeep,
-                                                                                      Map<Integer, List<SectionRectangle>> mapToMerge) {
-
-        mapToMerge.forEach((page, rectangle) -> mapToKeep.merge(page, rectangle, SectionGridCreatorService::concatLists));
-        return mapToKeep;
-    }
-
-
-    private static List<SectionRectangle> concatLists(List<SectionRectangle> l1, List<SectionRectangle> l2) {
-
-        return Stream.concat(l1.stream(), l2.stream()).toList();
-    }
-
-}

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

@@ -0,0 +1,26 @@
+package com.knecon.fforesight.service.layoutparser.processor.services;
+
+import java.util.List;
+
+import org.springframework.stereotype.Service;
+
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.SimplifiedSectionText;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Section;
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.SimplifiedText;
+
+@Service
+public class SimplifiedSectionTextService {
+
+    public SimplifiedText toSimplifiedText(Document document) {
+
+        List<SimplifiedSectionText> simplifiedSectionTexts = document.getMainSections().stream().map(this::toSimplifiedSectionText).toList();
+        return SimplifiedText.builder().numberOfPages(document.getNumberOfPages()).sectionTexts(simplifiedSectionTexts).build();
+    }
+
+
+    private SimplifiedSectionText toSimplifiedSectionText(Section section) {
+
+        return SimplifiedSectionText.builder().sectionNumber(section.getTreeId().get(0)).text(section.getTextBlock().getSearchText()).build();
+    }
+}

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

@@ -12,7 +12,6 @@ import java.util.Set;
 
 import org.springframework.stereotype.Service;
 
-import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
 import com.knecon.fforesight.service.layoutparser.processor.model.table.Cell;
@@ -66,6 +65,17 @@ public class TableExtractionService {
     };
 
 
+    public boolean contains(Cell cell, double x, double y, double w, double h) {
+
+        if (cell.isEmpty() || w <= 0 || h <= 0) {
+            return false;
+        }
+        double x0 = cell.getX();
+        double y0 = cell.getY();
+        return (x >= x0 - 2 && y >= y0 - 2 && (x + w) <= x0 + cell.getWidth() + 2 && (y + h) <= y0 + cell.getHeight() + 2);
+    }
+
+
     /**
      * Finds tables on a page and moves textblocks into cells of the found tables.
      * Note: This algorithm uses Pdf Coordinate System where {0,0} rotated with the page rotation.
@@ -79,16 +89,17 @@ public class TableExtractionService {
      * @param cleanRulings The lines used to build the table.
      * @param page         Page object that contains textblocks and statistics.
      */
-    public void extractTables(CleanRulings cleanRulings, ClassificationPage page, LayoutParsingType layoutParsingType) {
+    public void extractTables(CleanRulings cleanRulings, ClassificationPage page) {
 
-        List<Cell> cells = findCells(cleanRulings.getHorizontal(), cleanRulings.getVertical(), layoutParsingType);
+        List<Cell> cells = findCells(cleanRulings.getHorizontal(), cleanRulings.getVertical());
 
         List<TextPageBlock> toBeRemoved = new ArrayList<>();
 
         for (AbstractPageBlock abstractPageBlock : page.getTextBlocks()) {
             TextPageBlock textBlock = (TextPageBlock) abstractPageBlock;
             for (Cell cell : cells) {
-                if (cell.hasMinimumSize() && cell.intersects(textBlock.getPdfMinX(),
+                if (cell.hasMinimumSize() && contains(cell,
+                        textBlock.getPdfMinX(),
                         textBlock.getPdfMinY(),
                         textBlock.getPdfMaxX() - textBlock.getPdfMinX(),
                         textBlock.getPdfMaxY() - textBlock.getPdfMinY())) {
@@ -102,7 +113,7 @@ public class TableExtractionService {
         cells = new ArrayList<>(new HashSet<>(cells));
         DoubleComparisons.sort(cells, Rectangle.ILL_DEFINED_ORDER);
 
-        List<Rectangle> spreadsheetAreas = findSpreadsheetsFromCells(cells).stream().filter(r -> r.getWidth() > 0f && r.getHeight() > 0f).toList();
+        List<Rectangle> spreadsheetAreas = findSpreadsheetsFromCells(cells);
 
         List<TablePageBlock> tables = new ArrayList<>();
         for (Rectangle area : spreadsheetAreas) {
@@ -135,16 +146,14 @@ public class TableExtractionService {
     }
 
 
-    public List<Cell> findCells(List<Ruling> horizontalRulingLines, List<Ruling> verticalRulingLines, LayoutParsingType layoutParsingType) {
+    public List<Cell> findCells(List<Ruling> horizontalRulingLines, List<Ruling> verticalRulingLines) {
 
-        if (layoutParsingType.equals(LayoutParsingType.TAAS)) {
-            // TODO: breaks some tables, for example "1 Abamectin Prr.pdf" try to fix this upstream in RulingCleaningService
-            for (Ruling r : horizontalRulingLines) {
-                if (r.getX2() < r.getX1()) {
-                    double a = r.getX2();
-                    r.x2 = (float) r.getX1();
-                    r.x1 = (float) a;
-                }
+        // Fix for 211.pdf
+        for (Ruling r : horizontalRulingLines) {
+            if (r.getX2() < r.getX1()) {
+                double a = r.getX2();
+                r.x2 = (float) r.getX1();
+                r.x1 = (float) a;
             }
         }
 

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

@@ -1,74 +0,0 @@
-package com.knecon.fforesight.service.layoutparser.processor.services;
-
-import java.io.IOException;
-import java.io.InputStream;
-import java.util.Collection;
-import java.util.LinkedList;
-import java.util.List;
-import java.util.Map;
-import java.util.stream.Collectors;
-
-import org.apache.pdfbox.Loader;
-import org.apache.pdfbox.pdmodel.PDDocument;
-import org.apache.pdfbox.pdmodel.PDPage;
-import org.springframework.core.io.ClassPathResource;
-
-import com.knecon.fforesight.service.layoutparser.processor.model.PageContents;
-import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
-import com.knecon.fforesight.service.layoutparser.processor.services.parsing.PDFLinesTextStripper;
-import com.knecon.fforesight.service.layoutparser.processor.utils.RectangleTransformations;
-
-import lombok.experimental.UtilityClass;
-
-@UtilityClass
-public class TextPositionSequenceSorter {
-
-    public List<PageContents> getSortedTextPositionsWithPages(String filename) throws IOException {
-
-        List<PageContents> textPositionSequencesPerPage = new LinkedList<>();
-        try (InputStream inputStream = new ClassPathResource(filename).getInputStream()) {
-
-            try (PDDocument pdDocument = Loader.loadPDF(inputStream)) {
-
-                for (int pageNumber = 1; pageNumber < pdDocument.getNumberOfPages() + 1; pageNumber++) {
-
-                    PDFLinesTextStripper stripper = new PDFLinesTextStripper();
-                    PDPage pdPage = pdDocument.getPage(pageNumber - 1);
-                    stripper.setPageNumber(pageNumber);
-                    stripper.setSortByPosition(true);
-                    stripper.setStartPage(pageNumber);
-                    stripper.setEndPage(pageNumber);
-                    stripper.setPdpage(pdPage);
-                    stripper.getText(pdDocument);
-
-                    Map<Float, List<TextPositionSequence>> sortedTextPositionSequencesPerDir = stripper.getTextPositionSequences()
-                            .stream()
-                            .collect(Collectors.groupingBy(textPositionSequence -> textPositionSequence.getDir().getDegrees()));
-
-                    var sortedTextPositionSequences = sortByDirAccordingToPageRotation(sortedTextPositionSequencesPerDir, pdPage.getRotation());
-
-                    textPositionSequencesPerPage.add(new PageContents(sortedTextPositionSequences,
-                            RectangleTransformations.toRectangle2D(pdPage.getCropBox()),
-                            RectangleTransformations.toRectangle2D(pdPage.getMediaBox())));
-                }
-            }
-        }
-
-        return textPositionSequencesPerPage;
-    }
-
-
-    public List<TextPositionSequence> sortByDirAccordingToPageRotation(Map<Float, List<TextPositionSequence>> sortedTextPositionSequencesPerDir, int rotation) {
-
-        LinkedList<Float> sortedKeys = new LinkedList<>(sortedTextPositionSequencesPerDir.keySet().stream().sorted().toList());
-
-        for (int i = 0; i < sortedKeys.size(); i++) {
-            if (sortedKeys.get(i) < rotation) {
-                Float keyToSwap = sortedKeys.remove(i);
-                sortedKeys.addLast(keyToSwap);
-            }
-        }
-        return sortedKeys.stream().map(sortedTextPositionSequencesPerDir::get).flatMap(Collection::stream).toList();
-    }
-
-}

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

@@ -5,6 +5,9 @@ import static java.util.stream.Collectors.toSet;
 import java.util.ArrayList;
 import java.util.Comparator;
 import java.util.List;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+import java.util.stream.Collectors;
 
 import org.springframework.stereotype.Service;
 
@@ -23,6 +26,8 @@ public class DocuMineBlockificationService {
 
     static final float THRESHOLD = 1f;
 
+    Pattern pattern = Pattern.compile("^(\\d{1,2}\\.){1,3}\\d{1,2}\\.?\\s[0-9A-Za-z ()-]{2,50}", Pattern.CASE_INSENSITIVE);
+
 
     /**
      * This method is building blocks by expanding the minX/maxX and minY/maxY value on each word that is not split by the conditions.
@@ -43,7 +48,6 @@ public class DocuMineBlockificationService {
         float maxX = 0;
         float minY = 1000;
         float maxY = 0;
-
         TextPositionSequence prev = null;
 
         boolean wasSplitted = false;
@@ -57,10 +61,13 @@ public class DocuMineBlockificationService {
             boolean newLineAfterSplit = prev != null && word.getMinYDirAdj() != prev.getMinYDirAdj() && wasSplitted && splitX1 != word.getMinXDirAdj();
             boolean isSplitByRuling = isSplitByRuling(minX, minY, maxX, maxY, word, horizontalRulingLines, verticalRulingLines);
             boolean splitByDir = prev != null && !prev.getDir().equals(word.getDir());
-            boolean splitByOtherFontAndOtherY = prev != null && prev.getMaxYDirAdj() != word.getMaxYDirAdj() && (word.getFontStyle().contains("bold") && !prev.getFontStyle()
-                    .contains("bold") || prev.getFontStyle().contains("bold") && !word.getFontStyle().contains("bold"));
+            boolean splitByOtherFontAndOtherY = prev != null && Math.abs(prev.getMaxYDirAdj() - word.getMaxYDirAdj()) > word.getTextHeight() * 0.2 && (word.getFontStyle()
+                    .contains("bold") && !prev.getFontStyle().contains("bold") || prev.getFontStyle().contains("bold") && !word.getFontStyle().contains("bold"));
 
-            if (prev != null && (lineSeparation || startFromTop || splitByDir || isSplitByRuling || splitByOtherFontAndOtherY || negativeXGap)) {
+            Matcher matcher = pattern.matcher(chunkWords.stream().collect(Collectors.joining(" ")).toString());
+            boolean startsOnSameX = Math.abs(minX - word.getMinXDirAdj()) < 5 && matcher.matches();
+
+            if (prev != null && (lineSeparation || startFromTop || splitByDir || isSplitByRuling || splitByOtherFontAndOtherY || negativeXGap || startsOnSameX)) {
 
                 Orientation prevOrientation = null;
                 if (!chunkBlockList1.isEmpty()) {
@@ -231,3 +238,4 @@ public class DocuMineBlockificationService {
     }
 
 }
+

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

@@ -1,13 +1,8 @@
 package com.knecon.fforesight.service.layoutparser.processor.services.blockification;
 
-import java.util.ArrayList;
-import java.util.Iterator;
-import java.util.LinkedList;
-import java.util.List;
-import java.util.regex.Matcher;
-import java.util.regex.Pattern;
 
-import org.springframework.stereotype.Service;
+// TODO: figure out, why this fails the build
+// import static com.knecon.fforesight.service.layoutparser.processor.services.factory.SearchTextWithTextPositionFactory.HEIGHT_PADDING;
 
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
@@ -16,14 +11,25 @@ import com.knecon.fforesight.service.layoutparser.processor.model.table.Ruling;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
 import com.knecon.fforesight.service.layoutparser.processor.utils.RulingTextDirAdjustUtil;
+import org.springframework.stereotype.Service;
+
+import java.util.*;
+import java.util.regex.Matcher;
+import java.util.regex.Pattern;
+import java.util.stream.Stream;
 
 @Service
 @SuppressWarnings("all")
 public class TaasBlockificationService {
 
     private static final float THRESHOLD = 1f;
-    private static final float Y_GAP_SPLIT_HEIGHT_MODIFIER = 1.25f;
+    private static final float Y_GAP_SPLIT_HEIGHT_MODIFIER = 1.25f; // multiplied with text height
+    private static final float INTERSECTS_Y_THRESHOLD = 4;// 2 * HEIGHT_PADDING // This is exactly 2 times our position height padding. This is required to find boxes that are visually intersecting.
     private static final int X_GAP_SPLIT_CONSTANT = 50;
+    public static final int X_ALIGNMENT_THRESHOLD = 1;
+    public static final int NEGATIVE_X_GAP_THRESHOLD = -5;
+
+    private Pattern listIdentifier = Pattern.compile("^(?:(?:[1-9]|1\\d|20|[ivxlc]|[a-z])\\s*(?:[.)]))|\\uF0B7", Pattern.CASE_INSENSITIVE);
 
 
     /**
@@ -39,14 +45,28 @@ public class TaasBlockificationService {
     public ClassificationPage blockify(List<TextPositionSequence> textPositions, List<Ruling> horizontalRulingLines, List<Ruling> verticalRulingLines) {
 
         List<TextPageBlock> classificationTextBlocks = constructFineGranularTextPageBlocks(textPositions, horizontalRulingLines, verticalRulingLines);
-
-        classificationTextBlocks = mergeFineGranularTextPageBlocks(classificationTextBlocks);
+        classificationTextBlocks = mergeTextPageBlocksAligningX(classificationTextBlocks);
+        classificationTextBlocks = mergeIntersectingTextBlocksUntilConvergence(classificationTextBlocks);
 
         return new ClassificationPage(new ArrayList<>(classificationTextBlocks.stream().map(classificationTextBlock -> (AbstractPageBlock) classificationTextBlock).toList()));
     }
 
 
-    private List<TextPageBlock> mergeFineGranularTextPageBlocks(List<TextPageBlock> classificationTextBlocks) {
+    private List<TextPageBlock> mergeIntersectingTextBlocksUntilConvergence(List<TextPageBlock> classificationTextBlocks) {
+
+        int currentSize = classificationTextBlocks.size();
+        while (true) {
+            classificationTextBlocks = mergeTextPageBlocksAlmostIntersecting(classificationTextBlocks);
+            if (classificationTextBlocks.size() == currentSize) {
+                break;
+            }
+            currentSize = classificationTextBlocks.size();
+        }
+        return classificationTextBlocks;
+    }
+
+
+    private List<TextPageBlock> mergeTextPageBlocksAligningX(List<TextPageBlock> classificationTextBlocks) {
 
         if (classificationTextBlocks.isEmpty()) {
             return new ArrayList<>();
@@ -55,16 +75,29 @@ public class TaasBlockificationService {
         List<TextPageBlock> currentTextBlocksToMerge = new LinkedList<>();
         textBlocksToMerge.add(currentTextBlocksToMerge);
         TextPageBlock previousTextBlock = null;
+        Float lastLineGap = null;
         for (TextPageBlock currentTextBlock : classificationTextBlocks) {
             if (previousTextBlock == null) {
                 currentTextBlocksToMerge.add(currentTextBlock);
                 previousTextBlock = currentTextBlock;
                 continue;
             }
-            boolean alignsXRight = Math.abs(currentTextBlock.getPdfMaxX() - previousTextBlock.getPdfMaxX()) < 1;
-            boolean smallYGap = Math.abs(currentTextBlock.getPdfMaxY() - previousTextBlock.getPdfMinY()) < 5;
-            if (alignsXRight && smallYGap) {
+
+
+            Matcher listIdentifierPattern = listIdentifier.matcher(currentTextBlock.getText());
+            boolean isListIdentifier = listIdentifierPattern.find();
+
+            boolean yGap = Math.abs(currentTextBlock.getPdfMaxY() - previousTextBlock.getPdfMinY()) < previousTextBlock.getMostPopularWordHeight() * Y_GAP_SPLIT_HEIGHT_MODIFIER;
+
+            boolean sameFont = previousTextBlock.getMostPopularWordFont().equals(currentTextBlock.getMostPopularWordFont()) && previousTextBlock.getMostPopularWordFontSize() == currentTextBlock.getMostPopularWordFontSize();
+//            boolean yGap = previousTextBlock != null && currentTextBlock.getMinYDirAdj() - maxY > Math.min(word.getHeight(), prev.getHeight()) * Y_GAP_SPLIT_HEIGHT_MODIFIER;
+
+            boolean alignsXRight = Math.abs(currentTextBlock.getPdfMaxX() - previousTextBlock.getPdfMaxX()) < X_ALIGNMENT_THRESHOLD;
+            boolean alignsXLeft = Math.abs(currentTextBlock.getPdfMinX() - previousTextBlock.getPdfMinX()) < X_ALIGNMENT_THRESHOLD;
+//            boolean smallYGap = Math.abs(currentTextBlock.getPdfMaxY() - previousTextBlock.getPdfMinY()) < yGap;
+            if (yGap && sameFont && !isListIdentifier) {
                 currentTextBlocksToMerge.add(currentTextBlock);
+
             } else {
                 currentTextBlocksToMerge = new LinkedList<>();
                 currentTextBlocksToMerge.add(currentTextBlock);
@@ -76,6 +109,23 @@ public class TaasBlockificationService {
     }
 
 
+    private List<TextPageBlock> mergeTextPageBlocksAlmostIntersecting(List<TextPageBlock> textPageBlocks) {
+
+        Set<TextPageBlock> alreadyMerged = new HashSet<>();
+        List<List<TextPageBlock>> textBlocksToMerge = new LinkedList<>();
+        for (TextPageBlock textPageBlock : textPageBlocks) {
+            if (alreadyMerged.contains(textPageBlock)) {
+                continue;
+            }
+            alreadyMerged.add(textPageBlock);
+            textBlocksToMerge.add(Stream.concat(Stream.of(textPageBlock),
+                            textPageBlocks.stream().filter(textPageBlock2 -> textPageBlock.almostIntersects(textPageBlock2, INTERSECTS_Y_THRESHOLD, 0) && !alreadyMerged.contains(textPageBlock2)).peek(alreadyMerged::add))
+                    .toList());
+        }
+        return textBlocksToMerge.stream().map(TextPageBlock::merge).toList();
+    }
+
+
     private void assignOrientations(List<TextPageBlock> classificationTextBlocks) {
 
         Iterator<TextPageBlock> itty = classificationTextBlocks.iterator();
@@ -128,8 +178,8 @@ public class TaasBlockificationService {
 
 
     private List<TextPageBlock> constructFineGranularTextPageBlocks(List<TextPositionSequence> textPositions,
-                                                                                        List<Ruling> horizontalRulingLines,
-                                                                                        List<Ruling> verticalRulingLines) {
+                                                                    List<Ruling> horizontalRulingLines,
+                                                                    List<Ruling> verticalRulingLines) {
 
         int indexOnPage = 0;
         List<TextPositionSequence> wordClusterToCombine = new ArrayList<>();
@@ -138,18 +188,18 @@ public class TaasBlockificationService {
         float minX = 1000, maxX = 0, minY = 1000, maxY = 0;
         TextPositionSequence prev = null;
         // TODO: make static final constant
-        var listIdentitifier = Pattern.compile("\\b(?:[1-9]|1\\d|20|[ivxlc]|[a-z])\\s*(?:[.)])", Pattern.CASE_INSENSITIVE);
+
 
         boolean wasSplitted = false;
         Float splitX1 = null;
         for (TextPositionSequence word : textPositions) {
 
-            Matcher listIdentifierPattern = listIdentitifier.matcher(word.toString());
+            Matcher listIdentifierPattern = listIdentifier.matcher(word.toString());
 
             boolean yGap = prev != null && word.getMinYDirAdj() - maxY > Math.min(word.getHeight(), prev.getHeight()) * Y_GAP_SPLIT_HEIGHT_MODIFIER;
             boolean sameLine = prev != null && equalsWithThreshold(prev.getMinYDirAdj(), word.getMinYDirAdj());
             boolean positiveXGapInline = prev != null && maxX + X_GAP_SPLIT_CONSTANT < word.getMinXDirAdj() && sameLine;
-            boolean negativeXGap = prev != null && word.getMinXDirAdj() - minX < -5;
+            boolean negativeXGap = prev != null && word.getMinXDirAdj() - minX < NEGATIVE_X_GAP_THRESHOLD;
             boolean startFromTop = prev != null && word.getMinYDirAdj() < prev.getMinYDirAdj() - prev.getTextHeight();
             boolean newLineAfterSplit = prev != null && word.getMinYDirAdj() != prev.getMinYDirAdj() && wasSplitted && splitX1 != word.getMinXDirAdj();
             boolean splitByRuling = isSplitByRuling(minX, minY, maxX, maxY, word, horizontalRulingLines, verticalRulingLines);
@@ -164,7 +214,7 @@ public class TaasBlockificationService {
 
                 Orientation prevOrientation = null;
                 if (!classificationTextBlocks.isEmpty()) {
-                    prevOrientation = classificationTextBlocks.get(classificationTextBlocks.size() - 1).getOrientation();
+                    prevOrientation = classificationTextBlocks.get(classificationTextBlocks.size() - X_ALIGNMENT_THRESHOLD).getOrientation();
                 }
 
                 TextPageBlock classificationTextBlock = TextPageBlock.fromTextPositionSequences(wordClusterToCombine);

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

@@ -5,43 +5,36 @@ import java.util.Locale;
 import java.util.regex.Matcher;
 import java.util.regex.Pattern;
 
+import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 import org.springframework.stereotype.Service;
 
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
-import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
 import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.services.BodyTextFrameService;
 import com.knecon.fforesight.service.layoutparser.processor.utils.PositionUtils;
 
 import lombok.RequiredArgsConstructor;
 import lombok.extern.slf4j.Slf4j;
 
-
 @Slf4j
 @Service
 @RequiredArgsConstructor
 public class DocuMineClassificationService {
 
-    private final BodyTextFrameService bodyTextFrameService;
-    private static final Pattern pattern = Pattern.compile("^(\\d{1,1}\\.?){1,3}\\d{1,2}\\.?\\s[0-9A-Za-z\\[\\]\\-]{2,50}", Pattern.CASE_INSENSITIVE);
-    private static final Pattern pattern2 = Pattern.compile(".*\\d{4}$", Pattern.CASE_INSENSITIVE);
+    private static final Pattern pattern = Pattern.compile("^(\\d{1,2}\\.){1,3}\\d{1,2}\\.?\\s[0-9A-Za-z \\[\\]]{2,50}", Pattern.CASE_INSENSITIVE);
+    private static final Pattern pattern2 = Pattern.compile("\\p{L}{3,}", Pattern.CASE_INSENSITIVE);
     private static final Pattern pattern3 = Pattern.compile("^(\\d{1,1}\\.){1,3}\\d{1,2}\\.?\\s[a-z]{1,2}\\/[a-z]{1,2}.*");
 
 
     public void classifyDocument(ClassificationDocument document) {
 
-        Rectangle bodyTextFrame = bodyTextFrameService.calculateBodyTextFrame(document.getPages(), document.getFontSizeCounter(), false, LayoutParsingType.DOCUMINE);
-        Rectangle landscapeBodyTextFrame = bodyTextFrameService.calculateBodyTextFrame(document.getPages(), document.getFontSizeCounter(), true, LayoutParsingType.DOCUMINE);
         List<Float> headlineFontSizes = document.getFontSizeCounter().getHighterThanMostPopular();
 
         log.debug("Document FontSize counters are: {}", document.getFontSizeCounter().getCountPerValue());
 
         for (ClassificationPage page : document.getPages()) {
-            bodyTextFrameService.setBodyTextFrameAdjustedToPage(page, bodyTextFrame, landscapeBodyTextFrame);
             classifyPage(page, document, headlineFontSizes);
         }
     }
@@ -70,12 +63,16 @@ public class DocuMineClassificationService {
             textBlock.setClassification(PageBlockType.OTHER);
             return;
         }
-        if (PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
-                .getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular())) {
+        if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.HEADER)
+                || PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
+                .getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular())
+        ) {
             textBlock.setClassification(PageBlockType.HEADER);
 
-        } else if (PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
-                .getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular())) {
+        } else if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.FOOTER)
+                || PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
+                .getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular())
+        ) {
             textBlock.setClassification(PageBlockType.FOOTER);
         } else if (page.getPageNumber() == 1 && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock,
                 document.getTextHeightCounter().getMostPopular()) > 2.5 && textBlock.getHighestFontSize() > document.getFontSizeCounter().getMostPopular() || page.getTextBlocks()
@@ -115,4 +112,4 @@ public class DocuMineClassificationService {
         }
     }
 
-}
+}

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

@@ -3,16 +3,14 @@ package com.knecon.fforesight.service.layoutparser.processor.services.classifica
 import java.util.List;
 import java.util.regex.Pattern;
 
+import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 import org.springframework.stereotype.Service;
 
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
-import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
 import com.knecon.fforesight.service.layoutparser.processor.model.PageBlockType;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
-import com.knecon.fforesight.service.layoutparser.processor.services.BodyTextFrameService;
 import com.knecon.fforesight.service.layoutparser.processor.utils.PositionUtils;
 
 import lombok.RequiredArgsConstructor;
@@ -23,19 +21,14 @@ import lombok.extern.slf4j.Slf4j;
 @RequiredArgsConstructor
 public class RedactManagerClassificationService {
 
-    private final BodyTextFrameService bodyTextFrameService;
-
 
     public void classifyDocument(ClassificationDocument document) {
 
-        Rectangle bodyTextFrame = bodyTextFrameService.calculateBodyTextFrame(document.getPages(), document.getFontSizeCounter(), false, LayoutParsingType.REDACT_MANAGER);
-        Rectangle landscapeBodyTextFrame = bodyTextFrameService.calculateBodyTextFrame(document.getPages(), document.getFontSizeCounter(), true, LayoutParsingType.REDACT_MANAGER);
         List<Float> headlineFontSizes = document.getFontSizeCounter().getHighterThanMostPopular();
 
         log.debug("Document FontSize counters are: {}", document.getFontSizeCounter().getCountPerValue());
 
         for (ClassificationPage page : document.getPages()) {
-            bodyTextFrameService.setBodyTextFrameAdjustedToPage(page, bodyTextFrame, landscapeBodyTextFrame);
             classifyPage(page, document, headlineFontSizes);
         }
     }
@@ -59,11 +52,13 @@ public class RedactManagerClassificationService {
             textBlock.setClassification(PageBlockType.OTHER);
             return;
         }
-        if (PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
+        if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.HEADER)
+                || PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
                 .getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular())) {
             textBlock.setClassification(PageBlockType.HEADER);
 
-        } else if (PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
+        } else if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.FOOTER)
+                || PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation()) && (document.getFontSizeCounter()
                 .getMostPopular() == null || textBlock.getHighestFontSize() <= document.getFontSizeCounter().getMostPopular())) {
             textBlock.setClassification(PageBlockType.FOOTER);
         } else if (page.getPageNumber() == 1 && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock,

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

@@ -3,10 +3,9 @@ package com.knecon.fforesight.service.layoutparser.processor.services.classifica
 import java.util.List;
 import java.util.regex.Pattern;
 
+import com.knecon.fforesight.service.layoutparser.processor.utils.MarkedContentUtils;
 import org.springframework.stereotype.Service;
 
-import com.iqser.red.service.persistence.service.v1.api.shared.model.redactionlog.Rectangle;
-import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
 import com.knecon.fforesight.service.layoutparser.processor.model.AbstractPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationDocument;
 import com.knecon.fforesight.service.layoutparser.processor.model.ClassificationPage;
@@ -28,14 +27,13 @@ public class TaasClassificationService {
 
     public void classifyDocument(ClassificationDocument document) {
 
-        Rectangle bodyTextFrame = bodyTextFrameService.calculateBodyTextFrame(document.getPages(), document.getFontSizeCounter(), false, LayoutParsingType.TAAS);
-        Rectangle landscapeBodyTextFrame = bodyTextFrameService.calculateBodyTextFrame(document.getPages(), document.getFontSizeCounter(), true, LayoutParsingType.TAAS);
+
         List<Float> headlineFontSizes = document.getFontSizeCounter().getHighterThanMostPopular();
 
         log.debug("Document FontSize counters are: {}", document.getFontSizeCounter().getCountPerValue());
 
         for (ClassificationPage page : document.getPages()) {
-            bodyTextFrameService.setBodyTextFrameAdjustedToPage(page, bodyTextFrame, landscapeBodyTextFrame);
+
             classifyPage(page, document, headlineFontSizes);
         }
     }
@@ -59,9 +57,11 @@ public class TaasClassificationService {
             textBlock.setClassification(PageBlockType.OTHER);
             return;
         }
-        if (PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation())) {
+        if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.HEADER)
+                || PositionUtils.isOverBodyTextFrame(bodyTextFrame, textBlock, page.getRotation())) {
             textBlock.setClassification(PageBlockType.HEADER);
-        } else if (PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation())) {
+        } else if (MarkedContentUtils.intersects(textBlock, page.getMarkedContentBboxPerType(), MarkedContentUtils.FOOTER)
+                || PositionUtils.isUnderBodyTextFrame(bodyTextFrame, textBlock, page.getRotation())) {
             textBlock.setClassification(PageBlockType.FOOTER);
         } else if (page.getPageNumber() == 1 && (PositionUtils.getHeightDifferenceBetweenChunkWordAndDocumentWord(textBlock,
                 document.getTextHeightCounter().getMostPopular()) > 2.5 && textBlock.getHighestFontSize() > document.getFontSizeCounter().getMostPopular() || page.getTextBlocks()

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/factory/DocumentGraphFactory.java

@@ -64,7 +64,7 @@ public class DocumentGraphFactory {
 
     private void addSections(ClassificationDocument document, Context context) {
 
-        document.getSections().forEach(section -> SectionNodeFactory.addSection(null, section.getPageBlocks(), section.getImages(), context));
+        document.getSections().forEach(section -> SectionNodeFactory.addSection(null, section.getNonEmptyPageBlocks(), section.getImages(), context));
     }
 
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/factory/SearchTextWithTextPositionFactory.java

@@ -133,7 +133,7 @@ public class SearchTextWithTextPositionFactory {
 
     private static void addTextPositionWithFontType(RedTextPosition currentTextPosition, String fontType, List<Integer> fontTypePositions, int stringIdx) {
 
-        if (currentTextPosition.getFontName().toLowerCase(Locale.ROOT).contains(fontType)) {
+        if (currentTextPosition.getFontName() != null && currentTextPosition.getFontName().toLowerCase(Locale.ROOT).contains(fontType)) {
             fontTypePositions.add(stringIdx);
         }
     }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/factory/TableNodeFactory.java

@@ -32,7 +32,12 @@ public class TableNodeFactory {
         setPageNumberInCells(tablesToMerge);
         Set<Page> pages = tablesToMerge.stream().map(AbstractPageBlock::getPage).map(context::getPage).collect(Collectors.toSet());
         List<List<Cell>> mergedRows = tablesToMerge.stream().map(TablePageBlock::getRows).flatMap(Collection::stream).toList();
-        Table table = Table.builder().documentTree(context.getDocumentTree()).numberOfCols(mergedRows.get(0).size()).numberOfRows(mergedRows.size()).build();
+
+        Table table = Table.builder()
+                .documentTree(context.getDocumentTree())
+                .numberOfCols(mergedRows.isEmpty() ? 0 : mergedRows.get(0).size())
+                .numberOfRows(mergedRows.size())
+                .build();
 
         pages.forEach(page -> addTableToPage(page, parentNode, table));
 

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/mapper/PropertiesMapper.java

@@ -1,36 +1,27 @@
 package com.knecon.fforesight.service.layoutparser.processor.services.mapper;
 
 import java.awt.geom.Rectangle2D;
-import java.util.Arrays;
+import java.util.Collections;
 import java.util.HashMap;
-import java.util.List;
+import java.util.Locale;
 import java.util.Map;
 
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.DocumentStructure;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Image;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.ImageType;
-import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.TableCell;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Page;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Table;
-import com.knecon.fforesight.service.layoutparser.processor.utils.RectangleTransformations;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.TableCell;
 
 public class PropertiesMapper {
 
-    public static final String TRANSPARENT = "transparent";
-    public static final String IMAGE_TYPE = "imageType";
-    public static final String POSITION = "position";
-    public static final String ROW = "row";
-    public static final String COL = "col";
-    public static final String HEADER = "header";
-    public static final String B_BOX = "bBox";
-    public static final String NUMBER_OF_ROWS = "numberOfRows";
-    public static final String NUMBER_OF_COLS = "numberOfCols";
-
-
     public static Map<String, String> buildImageProperties(Image image) {
 
         Map<String, String> properties = new HashMap<>();
-        properties.put(IMAGE_TYPE, image.getImageType().toString());
-        properties.put(TRANSPARENT, String.valueOf(image.isTransparent()));
-        properties.put(POSITION, RectangleTransformations.toString(image.getPosition()));
+        properties.put(DocumentStructure.ImageProperties.IMAGE_TYPE, image.getImageType().toString());
+        properties.put(DocumentStructure.ImageProperties.TRANSPARENT, String.valueOf(image.isTransparent()));
+        properties.put(DocumentStructure.ImageProperties.POSITION, toString(image.getPosition()));
+        properties.put(DocumentStructure.ImageProperties.ID, image.getId());
         return properties;
     }
 
@@ -38,15 +29,15 @@ public class PropertiesMapper {
     public static Map<String, String> buildTableCellProperties(TableCell tableCell) {
 
         Map<String, String> properties = new HashMap<>();
-        properties.put(ROW, String.valueOf(tableCell.getRow()));
-        properties.put(COL, String.valueOf(tableCell.getCol()));
-        properties.put(HEADER, String.valueOf(tableCell.isHeader()));
+        properties.put(DocumentStructure.TableCellProperties.ROW, String.valueOf(tableCell.getRow()));
+        properties.put(DocumentStructure.TableCellProperties.COL, String.valueOf(tableCell.getCol()));
+        properties.put(DocumentStructure.TableCellProperties.HEADER, String.valueOf(tableCell.isHeader()));
 
         if (tableCell.getPages().size() > 1 || tableCell.getBBox().keySet().size() > 1) {
             throw new IllegalArgumentException("TableCell can only occur on a single page!");
         }
-        String bBoxString = RectangleTransformations.toString(tableCell.getBBox().get(tableCell.getPages().stream().findFirst().get()));
-        properties.put(B_BOX, bBoxString);
+        String bBoxString = toString(tableCell.getBBox().get(tableCell.getPages().stream().findFirst().get()));
+        properties.put(DocumentStructure.TableCellProperties.B_BOX, bBoxString);
 
         return properties;
     }
@@ -55,33 +46,33 @@ public class PropertiesMapper {
     public static Map<String, String> buildTableProperties(Table table) {
 
         Map<String, String> properties = new HashMap<>();
-        properties.put(NUMBER_OF_ROWS, String.valueOf(table.getNumberOfRows()));
-        properties.put(NUMBER_OF_COLS, String.valueOf(table.getNumberOfCols()));
+        properties.put(DocumentStructure.TableProperties.NUMBER_OF_ROWS, String.valueOf(table.getNumberOfRows()));
+        properties.put(DocumentStructure.TableProperties.NUMBER_OF_COLS, String.valueOf(table.getNumberOfCols()));
         return properties;
     }
 
 
     public static void parseImageProperties(Map<String, String> properties, Image.ImageBuilder builder) {
 
-        builder.imageType(parseImageType(properties.get(IMAGE_TYPE)));
-        builder.transparent(Boolean.parseBoolean(properties.get(TRANSPARENT)));
-        builder.position(parseRectangle2D(properties.get(POSITION)));
+        builder.imageType(parseImageType(properties.get(DocumentStructure.ImageProperties.IMAGE_TYPE)));
+        builder.transparent(Boolean.parseBoolean(properties.get(DocumentStructure.ImageProperties.TRANSPARENT)));
+        builder.position(DocumentStructure.parseRectangle2D(properties.get(DocumentStructure.ImageProperties.POSITION)));
     }
 
 
     public static void parseTableCellProperties(Map<String, String> properties, TableCell.TableCellBuilder builder) {
 
-        builder.row(Integer.parseInt(properties.get(ROW)));
-        builder.col(Integer.parseInt(properties.get(COL)));
-        builder.header(Boolean.parseBoolean(properties.get(HEADER)));
-        builder.bBox(parseRectangle2D(properties.get(B_BOX)));
+        builder.row(Integer.parseInt(properties.get(DocumentStructure.TableCellProperties.ROW)));
+        builder.col(Integer.parseInt(properties.get(DocumentStructure.TableCellProperties.COL)));
+        builder.header(Boolean.parseBoolean(properties.get(DocumentStructure.TableCellProperties.HEADER)));
+        builder.bBox(DocumentStructure.parseRectangle2D(properties.get(DocumentStructure.TableCellProperties.B_BOX)));
     }
 
 
     public static void parseTableProperties(Map<String, String> properties, Table.TableBuilder builder) {
 
-        builder.numberOfRows(Integer.parseInt(properties.get(NUMBER_OF_ROWS)));
-        builder.numberOfCols(Integer.parseInt(properties.get(NUMBER_OF_COLS)));
+        builder.numberOfRows(Integer.parseInt(properties.get(DocumentStructure.TableProperties.NUMBER_OF_ROWS)));
+        builder.numberOfCols(Integer.parseInt(properties.get(DocumentStructure.TableProperties.NUMBER_OF_COLS)));
     }
 
 
@@ -99,14 +90,15 @@ public class PropertiesMapper {
 
     public static String toString(Rectangle2D rectangle2D) {
 
-        return String.format("%f,%f,%f,%f", rectangle2D.getX(), rectangle2D.getY(), rectangle2D.getWidth(), rectangle2D.getHeight());
-    }
-
-
-    public static Rectangle2D parseRectangle2D(String bBox) {
-
-        List<Float> floats = Arrays.stream(bBox.split(",")).map(Float::parseFloat).toList();
-        return new Rectangle2D.Float(floats.get(0), floats.get(1), floats.get(2), floats.get(3));
+        return String.format(Locale.US,
+                "%f%s%f%s%f%s%f",
+                rectangle2D.getX(),
+                DocumentStructure.RECTANGLE_DELIMITER,
+                rectangle2D.getY(),
+                DocumentStructure.RECTANGLE_DELIMITER,
+                rectangle2D.getWidth(),
+                DocumentStructure.RECTANGLE_DELIMITER,
+                rectangle2D.getHeight());
     }
 
 }

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/parsing/LegacyPDFStreamEngine.java

@@ -92,28 +92,28 @@ public class LegacyPDFStreamEngine extends PDFStreamEngine {
      */
     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());
+        addOperator(new BeginText(this));
+        addOperator(new Concatenate(this));
+        addOperator(new DrawObject(this)); // special text version
+        addOperator(new EndText(this));
+        addOperator(new SetGraphicsStateParameters(this));
+        addOperator(new Save(this));
+        addOperator(new Restore(this));
+        addOperator(new NextLine(this));
+        addOperator(new SetCharSpacing(this));
+        addOperator(new MoveText(this));
+        addOperator(new MoveTextSetLeading(this));
+        addOperator(new SetFontAndSize(this));
+        addOperator(new ShowText(this));
+        addOperator(new ShowTextAdjusted(this));
+        addOperator(new SetTextLeading(this));
+        addOperator(new SetMatrix(this));
+        addOperator(new SetTextRenderingMode(this));
+        addOperator(new SetTextRise(this));
+        addOperator(new SetWordSpacing(this));
+        addOperator(new SetTextHorizontalScaling(this));
+        addOperator(new ShowTextLine(this));
+        addOperator(new ShowTextLineAndSpace(this));
 
         // load additional glyph list for Unicode mapping
         String path = "/org/apache/pdfbox/resources/glyphlist/additional.txt";

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/parsing/PDFLinesTextStripper.java

@@ -20,6 +20,8 @@ import org.apache.pdfbox.contentstream.operator.color.SetStrokingColorSpace;
 import org.apache.pdfbox.contentstream.operator.color.SetStrokingDeviceCMYKColor;
 import org.apache.pdfbox.contentstream.operator.color.SetStrokingDeviceGrayColor;
 import org.apache.pdfbox.contentstream.operator.color.SetStrokingDeviceRGBColor;
+import org.apache.pdfbox.contentstream.operator.markedcontent.BeginMarkedContentSequenceWithProperties;
+import org.apache.pdfbox.contentstream.operator.markedcontent.EndMarkedContentSequence;
 import org.apache.pdfbox.contentstream.operator.state.SetFlatness;
 import org.apache.pdfbox.contentstream.operator.state.SetLineCapStyle;
 import org.apache.pdfbox.contentstream.operator.state.SetLineDashPattern;
@@ -53,10 +55,6 @@ public class PDFLinesTextStripper extends PDFTextStripper {
     private final List<Ruling> graphicsPath = new ArrayList<>();
     @Setter
     protected PDPage pdpage;
-    private int minCharWidth;
-    private int maxCharWidth;
-    private int minCharHeight;
-    private int maxCharHeight;
 
     private float path_x;
     private float path_y;
@@ -68,26 +66,30 @@ public class PDFLinesTextStripper extends PDFTextStripper {
     public PDFLinesTextStripper() throws IOException {
 
         super();
-        this.addOperator(new SetStrokingColorSpace());
-        this.addOperator(new SetNonStrokingColorSpace());
-        this.addOperator(new SetLineDashPattern());
-        this.addOperator(new SetStrokingDeviceGrayColor());
-        this.addOperator(new SetNonStrokingDeviceGrayColor());
-        this.addOperator(new SetFlatness());
-        this.addOperator(new SetLineJoinStyle());
-        this.addOperator(new SetLineCapStyle());
-        this.addOperator(new SetStrokingDeviceCMYKColor());
-        this.addOperator(new SetNonStrokingDeviceCMYKColor());
-        this.addOperator(new SetLineMiterLimit());
-        this.addOperator(new SetStrokingDeviceRGBColor());
-        this.addOperator(new SetNonStrokingDeviceRGBColor());
-        this.addOperator(new SetRenderingIntent());
-        this.addOperator(new SetStrokingColor());
-        this.addOperator(new SetNonStrokingColor());
-        this.addOperator(new SetStrokingColorN());
-        this.addOperator(new SetNonStrokingColorN());
-        this.addOperator(new SetFontAndSize());
-        this.addOperator(new SetLineWidth());
+        this.addOperator(new SetStrokingColorSpace(this));
+        this.addOperator(new SetNonStrokingColorSpace(this));
+        this.addOperator(new SetLineDashPattern(this));
+        this.addOperator(new SetStrokingDeviceGrayColor(this));
+        this.addOperator(new SetNonStrokingDeviceGrayColor(this));
+        this.addOperator(new SetFlatness(this));
+        this.addOperator(new SetLineJoinStyle(this));
+        this.addOperator(new SetLineCapStyle(this));
+        this.addOperator(new SetStrokingDeviceCMYKColor(this));
+        this.addOperator(new SetNonStrokingDeviceCMYKColor(this));
+        this.addOperator(new SetLineMiterLimit(this));
+        this.addOperator(new SetStrokingDeviceRGBColor(this));
+        this.addOperator(new SetNonStrokingDeviceRGBColor(this));
+        this.addOperator(new SetRenderingIntent(this));
+        this.addOperator(new SetStrokingColor(this));
+        this.addOperator(new SetNonStrokingColor(this));
+        this.addOperator(new SetStrokingColorN(this));
+        this.addOperator(new SetNonStrokingColorN(this));
+        this.addOperator(new SetFontAndSize(this));
+        this.addOperator(new SetLineWidth(this));
+
+        addOperator(new BeginMarkedContentSequenceWithProperties(this));
+//        addOperator(new BeginMarkedContentSequence(this));
+        addOperator(new EndMarkedContentSequence(this));
     }
 
 
@@ -98,6 +100,7 @@ public class PDFLinesTextStripper extends PDFTextStripper {
 
         //move
         switch (operation) {
+
             case OperatorName.MOVE_TO:
                 if (arguments.size() == 2) {
                     Point2D.Float pos = transformPosition(floatValue(arguments.get(0)), floatValue(arguments.get(1)));
@@ -219,7 +222,11 @@ public class PDFLinesTextStripper extends PDFTextStripper {
             // This is a quick and dirt hack
             // Happens for file 216.pdf
             log.debug(e.getMessage());
-            return color.getComponents()[0] == 0 && color.getComponents()[1] == 0 && color.getComponents()[2] == 0 && color.getComponents()[1] == 1;
+            var result = true;
+            for (var component : color.getComponents()) {
+                result = result && component == 0;
+            }
+            return result;
         }
     }
 
@@ -238,33 +245,15 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                         .get(textPositionSequences.get(textPositionSequences.size() - 1).getTextPositions().size() - 1);
             }
 
-            int charWidth = (int) textPositions.get(i).getWidthDirAdj();
-            if (charWidth < minCharWidth) {
-                minCharWidth = charWidth;
-            }
-            if (charWidth > maxCharWidth) {
-                maxCharWidth = charWidth;
-            }
-
-            int charHeight = (int) textPositions.get(i).getHeightDir();
-            if (charHeight < minCharHeight) {
-                minCharHeight = charHeight;
-            }
-            if (charWidth > maxCharHeight) {
-                maxCharHeight = charHeight;
-            }
-
             if (i == 0 && (textPositions.get(i).getUnicode().equals(" ") || textPositions.get(i).getUnicode().equals("\u00A0") || textPositions.get(i).getUnicode().equals("\t"))) {
                 startIndex++;
                 continue;
             }
 
             // Strange but sometimes this is happening, for example: Metolachlor2.pdf
-            if (i > 0 && textPositions.get(i).getXDirAdj() < textPositions.get(i - 1).getXDirAdj()) {
+            if (checkIfCurrentPositionIsToTheRightOfPreviousPosition(i, textPositions)) {
                 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")))) {
+                if (checkIfSequenceContainsOnlyWhitespaces(sublist)) {
                     textPositionSequences.add(new TextPositionSequence(sublist, pageNumber, i == textPositions.size() - 1 && isParagraphStart));
                 }
                 startIndex = i;
@@ -272,9 +261,7 @@ public class PDFLinesTextStripper extends PDFTextStripper {
 
             if (textPositions.get(i).getRotation() == 0 && i > 0 && textPositions.get(i).getX() > textPositions.get(i - 1).getEndX() + 1) {
                 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")))) {
+                if (checkIfSequenceContainsOnlyWhitespaces(sublist)) {
                     textPositionSequences.add(new TextPositionSequence(sublist, pageNumber, i == textPositions.size() - 1 && isParagraphStart));
                 }
                 startIndex = i;
@@ -284,13 +271,10 @@ public class PDFLinesTextStripper extends PDFTextStripper {
                     .getUnicode()
                     .equals("\t")) && i <= textPositions.size() - 2) {
                 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")))) {
+                if (checkIfSequenceContainsOnlyWhitespaces(sublist)) {
 
                     // Remove false sequence ends (whitespaces)
-                    if (previous != null && sublist.get(0).getYDirAdj() == previous.getYDirAdj() && sublist.get(0)
-                            .getXDirAdj() - (previous.getXDirAdj() + previous.getWidthDirAdj()) < 0.01) {
+                    if (checkIfGapSizeBetweenCharactersSmallerThanMaximum(previous, sublist, 0.01f)) {
                         for (TextPosition t : sublist) {
                             textPositionSequences.get(textPositionSequences.size() - 1).add(t);
                         }
@@ -325,13 +309,29 @@ public class PDFLinesTextStripper extends PDFTextStripper {
     }
 
 
+    public boolean checkIfCurrentPositionIsToTheRightOfPreviousPosition(int i, List<TextPosition> textPositions) {
+
+        return i > 0 && textPositions.get(i).getXDirAdj() < textPositions.get(i - 1).getXDirAdj();
+    }
+
+
+    public boolean checkIfSequenceContainsOnlyWhitespaces(List<TextPosition> sublist) {
+
+        return !(sublist.isEmpty() || sublist.size() == 1 && (sublist.get(0).getUnicode().equals(" ") || sublist.get(0).getUnicode().equals("\u00A0") || sublist.get(0)
+                .getUnicode()
+                .equals("\t")));
+    }
+
+
+    public boolean checkIfGapSizeBetweenCharactersSmallerThanMaximum(RedTextPosition previous, List<TextPosition> sublist, float maximumGapSize) {
+
+        return previous != null && sublist.get(0).getYDirAdj() == previous.getYDirAdj() && sublist.get(0)
+                .getXDirAdj() - (previous.getXDirAdj() + previous.getWidthDirAdj()) < maximumGapSize;
+    }
+
     @Override
     public String getText(PDDocument doc) throws IOException {
 
-        minCharWidth = Integer.MAX_VALUE;
-        maxCharWidth = 0;
-        minCharHeight = Integer.MAX_VALUE;
-        maxCharHeight = 0;
         textPositionSequences.clear();
         rulings.clear();
         graphicsPath.clear();
@@ -344,3 +344,4 @@ public class PDFLinesTextStripper extends PDFTextStripper {
 
 }
 
+

layoutparser-service/layoutparser-service-processor/src/main/java/com/knecon/fforesight/service/layoutparser/processor/services/parsing/PDFTextStripper.java

@@ -25,32 +25,26 @@ import java.io.StringWriter;
 import java.io.Writer;
 import java.text.Bidi;
 import java.text.Normalizer;
-import java.util.ArrayList;
-import java.util.Collections;
-import java.util.Comparator;
-import java.util.HashMap;
-import java.util.Iterator;
-import java.util.LinkedList;
-import java.util.List;
-import java.util.Map;
-import java.util.SortedMap;
-import java.util.SortedSet;
-import java.util.StringTokenizer;
-import java.util.TreeMap;
-import java.util.TreeSet;
+import java.util.*;
 import java.util.regex.Pattern;
 
+import lombok.Getter;
 import org.apache.commons.logging.Log;
 import org.apache.commons.logging.LogFactory;
+import org.apache.pdfbox.cos.COSDictionary;
+import org.apache.pdfbox.cos.COSName;
 import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.pdmodel.PDPageTree;
 import org.apache.pdfbox.pdmodel.common.PDRectangle;
+import org.apache.pdfbox.pdmodel.documentinterchange.markedcontent.PDMarkedContent;
+import org.apache.pdfbox.pdmodel.graphics.PDXObject;
 import org.apache.pdfbox.pdmodel.interactive.documentnavigation.outline.PDOutlineItem;
 import org.apache.pdfbox.pdmodel.interactive.pagenavigation.PDThreadBead;
 import org.apache.pdfbox.text.TextPosition;
 import org.apache.pdfbox.text.TextPositionComparator;
-import org.apache.pdfbox.util.QuickSort;
+
+import com.knecon.fforesight.service.layoutparser.processor.utils.QuickSort;
 
 /**
  * This is just a copy except i only adjusted lines 594-607 cause this is a bug in Pdfbox.
@@ -63,6 +57,10 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
     private static float defaultDropThreshold = 2.5f;
     private static final boolean useCustomQuickSort;
 
+    @Getter
+    protected final List<PDMarkedContent> markedContents = new ArrayList<>();
+    protected final Deque<PDMarkedContent> currentMarkedContents = new ArrayDeque<>();
+
     private static final Log LOG = LogFactory.getLog(PDFTextStripper.class);
 
     // enable the ability to set the default indent/drop thresholds
@@ -196,6 +194,45 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
     }
 
 
+
+    public void beginMarkedContentSequence(COSName tag, COSDictionary properties)
+    {
+        PDMarkedContent markedContent = PDMarkedContent.create(tag, properties);
+        if (this.currentMarkedContents.isEmpty())
+        {
+            this.markedContents.add(markedContent);
+        }
+        else
+        {
+            PDMarkedContent currentMarkedContent =
+                    this.currentMarkedContents.peek();
+            if (currentMarkedContent != null)
+            {
+                currentMarkedContent.addMarkedContent(markedContent);
+            }
+        }
+        this.currentMarkedContents.push(markedContent);
+    }
+
+    @Override
+    public void endMarkedContentSequence()
+    {
+        if (!this.currentMarkedContents.isEmpty())
+        {
+            this.currentMarkedContents.pop();
+        }
+    }
+
+
+    public void xobject(PDXObject xobject)
+    {
+        if (!this.currentMarkedContents.isEmpty())
+        {
+            this.currentMarkedContents.peek().addXObject(xobject);
+        }
+    }
+
+
     /**
      * This will return the text of a document. See writeText. <br>
      * NOTE: The document must not be encrypted when coming into this method.
@@ -877,7 +914,12 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
                     textList.add(text);
                 }
             }
+            if (!this.currentMarkedContents.isEmpty())
+            {
+                this.currentMarkedContents.peek().addText(text);
+            }
         }
+
     }
 
 
@@ -1669,7 +1711,7 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
         int numberOfStrings = line.size();
         for (int i = 0; i < numberOfStrings; i++) {
             WordWithTextPositions word = line.get(i);
-            word.getTextPositions().sort(Comparator.comparing(TextPosition::getX));
+            word.getTextPositions().sort(Comparator.comparing(TextPosition::getXDirAdj));
             writeString(word.getText(), word.getTextPositions(), isParagraphEnd && i == numberOfStrings - 1);
             if (i < numberOfStrings - 1) {
                 writeWordSeparator();
@@ -2103,6 +2145,7 @@ public class PDFTextStripper extends LegacyPDFStreamEngine {
             this.isHangingIndent = true;
         }
 
+
     }
 
 }

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

@@ -0,0 +1,282 @@
+package com.knecon.fforesight.service.layoutparser.processor.services.visualization;
+
+import static com.knecon.fforesight.service.layoutparser.processor.services.visualization.ViewerDocumentService.LINE_WIDTH;
+
+import java.awt.Color;
+import java.awt.geom.Line2D;
+import java.awt.geom.Point2D;
+import java.awt.geom.Rectangle2D;
+import java.awt.geom.RectangularShape;
+import java.util.ArrayList;
+import java.util.Collections;
+import java.util.Comparator;
+import java.util.List;
+import java.util.Map;
+import java.util.Optional;
+import java.util.stream.Collectors;
+import java.util.stream.Stream;
+
+import org.springframework.stereotype.Service;
+
+import com.knecon.fforesight.service.layoutparser.internal.api.data.redaction.NodeType;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Page;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SemanticNode;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Table;
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.TableCell;
+import com.knecon.fforesight.service.layoutparser.processor.model.visualization.ColoredLine;
+import com.knecon.fforesight.service.layoutparser.processor.model.visualization.ColoredRectangle;
+import com.knecon.fforesight.service.layoutparser.processor.model.visualization.LayoutGrid;
+import com.knecon.fforesight.service.layoutparser.processor.model.visualization.PlacedText;
+import com.knecon.fforesight.service.layoutparser.processor.utils.RectangleTransformations;
+
+@Service
+public class LayoutGridService {
+
+    private static final Color INNER_LINES_COLOR = new Color(255, 175, 175);
+    private static final Color PARAGRAPH_COLOR = new Color(70, 130, 180);
+    public static final Color TABLE_COLOR = new Color(102, 205, 170);
+    public static final Color SECTION_COLOR = new Color(50, 50, 50);
+    public static final Color HEADLINE_COLOR = new Color(162, 56, 56);
+    public static final Color HEADER_COLOR = new Color(171, 131, 6);
+    public static final Color IMAGE_COLOR = new Color(253, 63, 146);
+
+
+    public LayoutGrid createLayoutGrid(Document document) {
+
+        LayoutGrid layoutGrid = new LayoutGrid(document.getNumberOfPages());
+        document.streamAllSubNodes().forEach(semanticNode -> {
+            Color color = switch (semanticNode.getType()) {
+                case PARAGRAPH -> PARAGRAPH_COLOR;
+                case TABLE -> TABLE_COLOR;
+                case SECTION -> SECTION_COLOR;
+                case HEADLINE -> HEADLINE_COLOR;
+                case HEADER, FOOTER -> HEADER_COLOR;
+                case IMAGE -> IMAGE_COLOR;
+                default -> null;
+            };
+            if (isNotSectionOrTableCellOrDocument(semanticNode)) {
+                addAsRectangle(semanticNode, layoutGrid, color);
+            }
+            if (semanticNode.getType().equals(NodeType.SECTION)) {
+                addSection(semanticNode, layoutGrid, color);
+            }
+            if (semanticNode.getType().equals(NodeType.TABLE)) {
+                Table table = (Table) semanticNode;
+                addInnerTableLines(table, layoutGrid);
+            }
+        });
+
+        return layoutGrid;
+    }
+
+
+    private void addInnerTableLines(Table table, LayoutGrid layoutGrid) {
+
+        if (table.getNumberOfCols() < 1 || table.getNumberOfRows() < 1) {
+            return;
+        }
+        for (Page page : table.getPages()) {
+            Optional<Integer> optionalFirstRowOnPage = table.streamCol(0).filter(tableCell -> tableCell.isOnPage(page.getNumber())).map(TableCell::getRow).findFirst();
+            if (optionalFirstRowOnPage.isEmpty()) {
+                continue;
+            }
+            int firstRowOnPage = optionalFirstRowOnPage.get();
+            Stream<Double> xStream = switch (page.getRotation()) {
+                case 90 -> streamBBoxOfCellsOnPage(table.streamCol(0), page).map(RectangularShape::getMinX);
+                case 180 -> streamBBoxOfCellsOnPage(table.streamRow(firstRowOnPage), page).map(RectangularShape::getMaxX);
+                case 270 -> streamBBoxOfCellsOnPage(table.streamCol(0), page).map(RectangularShape::getMaxX);
+                default -> streamBBoxOfCellsOnPage(table.streamRow(firstRowOnPage), page).map(RectangularShape::getMinX);
+            };
+            List<Double> xs = xStream.collect(Collectors.toList());
+            xs.remove(0);
+            Stream<Double> yStream = switch (page.getRotation()) {
+                case 90 -> streamBBoxOfCellsOnPage(table.streamRow(firstRowOnPage), page).map(RectangularShape::getMinY);
+                case 180 -> streamBBoxOfCellsOnPage(table.streamCol(0), page).map(RectangularShape::getMinY);
+                case 270 -> streamBBoxOfCellsOnPage(table.streamRow(firstRowOnPage), page).map(RectangularShape::getMaxY);
+                default -> streamBBoxOfCellsOnPage(table.streamCol(0), page).map(RectangularShape::getMaxY);
+            };
+            List<Double> ys = yStream.collect(Collectors.toList());
+            ys.remove(0);
+
+            Rectangle2D tableBBox = table.getBBox().get(table.getFirstPage());
+            List<ColoredLine> coloredLines = layoutGrid.getVisualizationsPerPages().get(page.getNumber() - 1).getColoredLines();
+            xs.forEach(x -> {
+                Line2D line = new Line2D.Double(new Point2D.Double(x, tableBBox.getMaxY()), new Point2D.Double(x, tableBBox.getMinY()));
+                coloredLines.add(new ColoredLine(line, INNER_LINES_COLOR));
+            });
+            ys.forEach(y -> {
+                Line2D line = new Line2D.Double(new Point2D.Double(tableBBox.getMinX(), y), new Point2D.Double(tableBBox.getMaxX(), y));
+                coloredLines.add(new ColoredLine(line, INNER_LINES_COLOR));
+            });
+        }
+    }
+
+
+    private static Stream<Rectangle2D> streamBBoxOfCellsOnPage(Stream<TableCell> table, Page page) {
+
+        return table.filter(tableCell -> tableCell.isOnPage(page.getNumber())).map(TableCell::getBBox).map(bBoxMap -> bBoxMap.get(page));
+    }
+
+
+    private void addSection(SemanticNode semanticNode, LayoutGrid layoutGrid, Color color) {
+
+        Map<Page, Rectangle2D> bBoxMap = semanticNode.getBBox();
+        List<SemanticNode> subSections = semanticNode.streamAllSubNodesOfType(NodeType.SECTION).toList();
+        Page firstPage = semanticNode.getFirstPage();
+        if (!subSections.isEmpty()) {
+            addPlacedText(firstPage, bBoxMap.get(firstPage), buildTreeIdString(semanticNode), layoutGrid);
+        } else {
+            bBoxMap.forEach(((page, textBBox) -> addPlacedText(page, textBBox, buildTreeIdString(semanticNode), layoutGrid)));
+        }
+        if (bBoxMap.values().size() == 1) {
+            Rectangle2D r = RectangleTransformations.pad(bBoxMap.get(firstPage), LINE_WIDTH, LINE_WIDTH);
+            List<ColoredLine> coloredLines = layoutGrid.getVisualizationsPerPages().get(firstPage.getNumber() - 1).getColoredLines();
+            List<Line2D> lines = createLinesFromRectangle(r, firstPage.getRotation());
+            // add string to top line
+            var firstLine = lines.remove(0);
+            coloredLines.add(new ColoredLine(firstLine, color));
+            for (Line2D line : lines) {
+                coloredLines.add(new ColoredLine(line, color));
+            }
+            return;
+        }
+        List<Page> pagesInOrder = bBoxMap.keySet().stream().sorted(Comparator.comparingInt(Page::getNumber)).collect(Collectors.toList());
+        pagesInOrder.remove(0);
+        addLinesForFirstPageOfSection(semanticNode, color, firstPage, layoutGrid);
+        var lastPage = pagesInOrder.remove(pagesInOrder.size() - 1);
+        addLinesForLastPageOfSection(semanticNode, color, lastPage, layoutGrid);
+        for (Page middlePage : pagesInOrder) {
+            addLinesForMiddlePageOfSection(semanticNode, color, middlePage, layoutGrid);
+        }
+    }
+
+
+    private void addPlacedText(Page page, Rectangle2D textBBox, String s, LayoutGrid layoutGrid) {
+
+        Point2D.Float upperLeftCorner = switch (page.getRotation()) {
+            case 90 -> new Point2D.Float((float) (textBBox.getMinX()), (float) textBBox.getMinY());
+            case 180 -> new Point2D.Float((float) (textBBox.getMaxX()), (float) textBBox.getMinY());
+            case 270 -> new Point2D.Float((float) (textBBox.getMaxX()), (float) textBBox.getMaxY());
+            default -> new Point2D.Float((float) (textBBox.getMinX()), (float) textBBox.getMaxY());
+        };
+        var placedTexts = layoutGrid.getVisualizationsPerPages().get(page.getNumber() - 1).getPlacedTexts();
+        placedTexts.add(new PlacedText(s, upperLeftCorner));
+    }
+
+
+    private void addLinesForMiddlePageOfSection(SemanticNode semanticNode, Color color, Page middlePage, LayoutGrid layoutGrid) {
+
+        List<ColoredLine> coloredLines = layoutGrid.getVisualizationsPerPages().get(middlePage.getNumber() - 1).getColoredLines();
+        Rectangle2D r = RectangleTransformations.pad(semanticNode.getBBox().get(middlePage), LINE_WIDTH, LINE_WIDTH);
+        var midPageLines = createLinesFromRectangle(r, middlePage.getRotation());
+        // remove top line
+        midPageLines.remove(0);
+        // remove top line
+        midPageLines.remove(1);
+        // add string to left line
+        var leftLine = midPageLines.remove(1);
+        coloredLines.add(new ColoredLine(leftLine, color));
+        for (Line2D line : midPageLines) {
+            coloredLines.add(new ColoredLine(line, color));
+        }
+    }
+
+
+    private void addLinesForLastPageOfSection(SemanticNode semanticNode, Color color, Page lastPage, LayoutGrid layoutGrid) {
+
+        List<ColoredLine> coloredLines = layoutGrid.getVisualizationsPerPages().get(lastPage.getNumber() - 1).getColoredLines();
+        Rectangle2D r = RectangleTransformations.pad(semanticNode.getBBox().get(lastPage), LINE_WIDTH, LINE_WIDTH);
+        var lastPageLines = createLinesFromRectangle(r, lastPage.getRotation());
+        // remove top line
+        lastPageLines.remove(0);
+        // add string to left line
+        var leftLine = lastPageLines.remove(2);
+        coloredLines.add(new ColoredLine(leftLine, color));
+        for (Line2D line : lastPageLines) {
+            coloredLines.add(new ColoredLine(line, color));
+        }
+    }
+
+
+    private void addLinesForFirstPageOfSection(SemanticNode semanticNode, Color color, Page firstPage, LayoutGrid layoutGrid) {
+
+        List<ColoredLine> coloredLines = layoutGrid.getVisualizationsPerPages().get(firstPage.getNumber() - 1).getColoredLines();
+        Rectangle2D r = RectangleTransformations.pad(semanticNode.getBBox().get(firstPage), LINE_WIDTH, LINE_WIDTH);
+        var firstPageLines = createLinesFromRectangle(r, firstPage.getRotation());
+        // remove bottom line
+        firstPageLines.remove(2);
+        // add string to top line
+        var firstLine = firstPageLines.remove(0);
+        coloredLines.add(new ColoredLine(firstLine, color));
+        for (Line2D line : firstPageLines) {
+            coloredLines.add(new ColoredLine(line, color));
+        }
+    }
+
+
+    private String buildTreeIdString(SemanticNode semanticNode) {
+
+        return semanticNode.getTreeId().stream().map(Object::toString).collect(Collectors.joining("."));
+    }
+
+
+    /*
+    A __________________ B
+     |                  |
+     |                  |
+     |                  |
+     |                  |
+    D|__________________| C
+    The returned List are the lines [AB, BC, DC, AD]
+    The List is reordered, such that the order of the returned lines are always as viewed on the page.
+     */
+    private List<Line2D> createLinesFromRectangle(Rectangle2D r, int pageRotation) {
+        // +0.5 to join the lines
+        List<Line2D> lines = new ArrayList<>(4);
+        float lineWidthCorrection = LINE_WIDTH * 0.5f;
+        Point2D.Float a = new Point2D.Float((float) r.getMinX(), (float) r.getMaxY());
+        Point2D.Float a1 = new Point2D.Float((float) r.getMinX() - lineWidthCorrection, (float) r.getMaxY());
+        Point2D.Float b = new Point2D.Float((float) r.getMaxX(), (float) r.getMaxY());
+        Point2D.Float b1 = new Point2D.Float((float) r.getMaxX() + lineWidthCorrection, (float) r.getMaxY());
+        Point2D.Float c = new Point2D.Float((float) r.getMaxX(), (float) r.getMinY());
+        Point2D.Float c1 = new Point2D.Float((float) r.getMaxX() + lineWidthCorrection, (float) r.getMinY());
+        Point2D.Float d = new Point2D.Float((float) r.getMinX(), (float) r.getMinY());
+        Point2D.Float d1 = new Point2D.Float((float) r.getMinX() - lineWidthCorrection, (float) r.getMinY());
+        lines.add(new Line2D.Float(a1, b1));
+        lines.add(new Line2D.Float(b, c));
+        lines.add(new Line2D.Float(d1, c1));
+        lines.add(new Line2D.Float(a, d));
+
+        return switch (pageRotation) {
+            case 90 -> {
+                Collections.rotate(lines, 1);
+                yield lines;
+            }
+            case 180 -> {
+                Collections.rotate(lines, 2);
+                yield lines;
+            }
+            case 270 -> {
+                Collections.rotate(lines, 3);
+                yield lines;
+            }
+
+            default -> lines;
+        };
+    }
+
+
+    private static boolean isNotSectionOrTableCellOrDocument(SemanticNode semanticNode) {
+
+        return !(semanticNode.getType().equals(NodeType.DOCUMENT) || semanticNode.getType().equals(NodeType.SECTION) || semanticNode.getType().equals(NodeType.TABLE_CELL));
+    }
+
+
+    private void addAsRectangle(SemanticNode semanticNode, LayoutGrid layoutGrid, Color color) {
+
+        semanticNode.getBBox()
+                .forEach((page, textBBox) -> layoutGrid.getVisualizationsPerPages().get(page.getNumber() - 1).getColoredRectangles().add(new ColoredRectangle(textBBox, color)));
+    }
+
+}

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

@@ -0,0 +1,172 @@
+package com.knecon.fforesight.service.layoutparser.processor.services.visualization;
+
+import java.awt.geom.AffineTransform;
+import java.awt.geom.Rectangle2D;
+import java.io.IOException;
+import java.io.OutputStream;
+
+import org.apache.pdfbox.cos.COSName;
+import org.apache.pdfbox.pdfwriter.compress.CompressParameters;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDDocumentCatalog;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.PDPageContentStream;
+import org.apache.pdfbox.pdmodel.PDResources;
+import org.apache.pdfbox.pdmodel.font.PDFont;
+import org.apache.pdfbox.pdmodel.font.PDType1Font;
+import org.apache.pdfbox.pdmodel.font.Standard14Fonts;
+import org.apache.pdfbox.pdmodel.graphics.optionalcontent.PDOptionalContentGroup;
+import org.apache.pdfbox.pdmodel.graphics.optionalcontent.PDOptionalContentProperties;
+import org.apache.pdfbox.pdmodel.graphics.state.PDExtendedGraphicsState;
+import org.apache.pdfbox.util.Matrix;
+import org.springframework.stereotype.Service;
+
+import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
+import com.knecon.fforesight.service.layoutparser.processor.model.visualization.ColoredLine;
+import com.knecon.fforesight.service.layoutparser.processor.model.visualization.ColoredRectangle;
+import com.knecon.fforesight.service.layoutparser.processor.model.visualization.FilledRectangle;
+import com.knecon.fforesight.service.layoutparser.processor.model.visualization.LayoutGrid;
+import com.knecon.fforesight.service.layoutparser.processor.model.visualization.PlacedText;
+import com.knecon.fforesight.service.layoutparser.processor.model.visualization.VisualizationsOnPage;
+
+import lombok.RequiredArgsConstructor;
+import lombok.SneakyThrows;
+import lombok.extern.slf4j.Slf4j;
+
+@Slf4j
+@Service
+@RequiredArgsConstructor
+public class ViewerDocumentService {
+
+    private static final String LAYER_NAME = "Layout grid";
+    private static final int FONT_SIZE = 10;
+    public static final float LINE_WIDTH = 1f;
+
+    private final LayoutGridService layoutGridService;
+
+
+    @SneakyThrows
+    public void createViewerDocument(PDDocument pdDocument, Document document, OutputStream outputStream, boolean layerVisibilityDefaultValue) {
+
+        LayoutGrid layoutGrid = layoutGridService.createLayoutGrid(document);
+        // PDDocument.save() is very slow, since it actually traverses the entire pdf and writes a new one.
+        // If we collect all COSDictionaries we changed and tell it explicitly to only add the changed ones by using saveIncremental it's very fast.
+        PDOptionalContentGroup layer = addLayerToDocument(pdDocument, layerVisibilityDefaultValue);
+        PDFont font = new PDType1Font(Standard14Fonts.FontName.HELVETICA);
+
+        for (int pageNumber = 0; pageNumber < pdDocument.getNumberOfPages(); pageNumber++) {
+            PDPage pdPage = pdDocument.getPage(pageNumber);
+
+            AffineTransform textDeRotationMatrix = getTextDeRotationTransform(pdPage);
+            addLayerToPageRessources(pdPage);
+
+            // We need to save the graphics state before, such that our appended content cannot be affected by previous content streams with side effects,
+            // e.g. not escaped matrix transformations.
+            escapePreviousContents(pdDocument, pdPage);
+
+            VisualizationsOnPage visualizationsOnPage = layoutGrid.getVisualizationsPerPages()
+                    .get(pageNumber);
+            assert pageNumber == visualizationsOnPage.getPageNumber();
+            // We need to append to the content stream, otherwise the content could be overlapped by following content.
+            try (var contentStream = new PDPageContentStream(pdDocument, pdPage, PDPageContentStream.AppendMode.APPEND, true)) {
+
+                contentStream.beginMarkedContent(COSName.OC, layer);
+                contentStream.saveGraphicsState();
+
+                contentStream.setLineWidth(LINE_WIDTH);
+                for (ColoredLine coloredLine : visualizationsOnPage.getColoredLines()) {
+                    contentStream.setStrokingColor(coloredLine.color());
+                    contentStream.moveTo((float) coloredLine.line().getX1(), (float) coloredLine.line().getY1());
+                    contentStream.lineTo((float) coloredLine.line().getX2(), (float) coloredLine.line().getY2());
+                    contentStream.stroke();
+                }
+                for (ColoredRectangle coloredRectangle : visualizationsOnPage.getColoredRectangles()) {
+                    contentStream.setStrokingColor(coloredRectangle.color());
+                    Rectangle2D r = coloredRectangle.rectangle2D();
+                    contentStream.addRect((float) r.getX(), (float) r.getY(), (float) r.getWidth(), (float) r.getHeight());
+                    contentStream.stroke();
+                }
+                for (FilledRectangle filledRectangle : visualizationsOnPage.getFilledRectangles()) {
+                    contentStream.setNonStrokingColor(filledRectangle.color());
+                    PDExtendedGraphicsState graphicsState = new PDExtendedGraphicsState();
+                    graphicsState.setNonStrokingAlphaConstant(filledRectangle.alpha());
+                    contentStream.setGraphicsStateParameters(graphicsState);
+                    Rectangle2D r = filledRectangle.rectangle2D();
+                    contentStream.addRect((float) r.getX(), (float) r.getY(), (float) r.getWidth(), (float) r.getHeight());
+                    contentStream.fill();
+                }
+                for (PlacedText placedText : visualizationsOnPage.getPlacedTexts()) {
+                    contentStream.setFont(font, FONT_SIZE);
+                    contentStream.beginText();
+                    Matrix textMatrix = new Matrix((float) textDeRotationMatrix.getScaleX(),
+                                                   (float) textDeRotationMatrix.getShearX(),
+                                                   (float) textDeRotationMatrix.getShearY(),
+                                                   (float) textDeRotationMatrix.getScaleY(),
+                                                   (float) placedText.lineStart().getX(),
+                                                   (float) placedText.lineStart().getY());
+                    textMatrix.translate(-((font.getStringWidth(placedText.text()) / 1000) * FONT_SIZE + (2 * LINE_WIDTH) + 4), -FONT_SIZE);
+                    contentStream.setTextMatrix(textMatrix);
+                    contentStream.showText(placedText.text());
+                    contentStream.endText();
+                }
+                contentStream.restoreGraphicsState();
+                contentStream.endMarkedContent();
+            }
+        }
+//        dictionariesToUpdate.add(pdDocument.getDocument().getTrailer());
+        pdDocument.save(outputStream, CompressParameters.NO_COMPRESSION);
+    }
+
+
+    private static void addLayerToPageRessources(PDPage pdPage) {
+
+        PDResources resources = pdPage.getResources();
+        if (resources == null) {
+            resources = new PDResources();
+            pdPage.setResources(resources);
+        }
+    }
+
+
+    private static void escapePreviousContents(PDDocument pdDocument, PDPage pdPage) throws IOException {
+
+        try (var contentStream = new PDPageContentStream(pdDocument, pdPage, PDPageContentStream.AppendMode.PREPEND, false)) {
+            contentStream.saveGraphicsState();
+        }
+        try (var contentStream = new PDPageContentStream(pdDocument, pdPage, PDPageContentStream.AppendMode.APPEND, false)) {
+            contentStream.restoreGraphicsState();
+        }
+    }
+
+
+    private static PDOptionalContentGroup addLayerToDocument(PDDocument pdDocument, boolean layerVisibilityDefaultValue) {
+
+        PDDocumentCatalog catalog = pdDocument.getDocumentCatalog();
+        PDOptionalContentProperties ocprops = catalog.getOCProperties();
+        if (ocprops == null) {
+            ocprops = new PDOptionalContentProperties();
+            catalog.setOCProperties(ocprops);
+        }
+        PDOptionalContentGroup layer = null;
+        if (ocprops.hasGroup(LAYER_NAME)) {
+            layer = ocprops.getGroup(LAYER_NAME);
+        } else {
+            layer = new PDOptionalContentGroup(LAYER_NAME);
+            ocprops.addGroup(layer);
+        }
+        ocprops.setGroupEnabled(layer, layerVisibilityDefaultValue);
+        return layer;
+    }
+
+
+    private static AffineTransform getTextDeRotationTransform(PDPage page) {
+
+        return AffineTransform.getQuadrantRotateInstance(switch (page.getRotation()) {
+            case 90 -> 3;
+            case 180 -> 2;
+            case 270 -> 1;
+            default -> 0;
+        });
+    }
+
+}

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

@@ -0,0 +1,56 @@
+package com.knecon.fforesight.service.layoutparser.processor.utils;
+
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
+import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;
+import lombok.experimental.UtilityClass;
+import org.apache.pdfbox.cos.COSName;
+import org.apache.pdfbox.pdmodel.documentinterchange.markedcontent.PDMarkedContent;
+import org.apache.pdfbox.text.TextPosition;
+
+import java.awt.geom.Rectangle2D;
+import java.util.Collection;
+import java.util.Collections;
+import java.util.List;
+import java.util.Map;
+import java.util.stream.Collectors;
+
+@UtilityClass
+public class MarkedContentUtils {
+
+    public static final String HEADER = "Header";
+    public static final String FOOTER = "Footer";
+
+    public List<Rectangle2D> getMarkedContentBboxPerLine(List<PDMarkedContent> markedContents, String subtype) {
+
+        if (markedContents == null) {
+            return Collections.emptyList();
+        }
+
+        var markedContentByYPosition = markedContents.stream()
+                .filter(m -> m.getTag().equals("Artifact"))
+                .filter(m -> m.getProperties() != null)
+                .filter(m -> m.getProperties().getItem("Subtype") != null)
+                .filter(m -> ((COSName) m.getProperties().getItem("Subtype")).getName().equals(subtype))
+                .map(PDMarkedContent::getContents).flatMap(Collection::stream)
+                .filter(t -> t instanceof TextPosition)
+                .map(t -> (TextPosition) t)
+                .filter(t -> !t.getUnicode().equals(" "))
+                .collect(Collectors.groupingBy(TextPosition::getY));
+
+        if (markedContentByYPosition.isEmpty()) {
+            return Collections.emptyList();
+        }
+
+        return markedContentByYPosition.values().stream()
+                .map(textPositions -> new TextPositionSequence(textPositions.stream()
+                        .toList(), 0, true)
+                        .getRectangle())
+                .map(t -> new Rectangle2D.Float(t.getTopLeft().getX(), t.getTopLeft().getY() - Math.abs(t.getHeight()), t.getWidth(), Math.abs(t.getHeight()))).collect(Collectors.toList());
+    }
+
+
+    public boolean intersects(TextPageBlock textBlock, Map<String, List<Rectangle2D>> markedContentBboxPerType, String type) {
+        return markedContentBboxPerType.get(type) != null && markedContentBboxPerType.get(type).stream().anyMatch(rectangle -> rectangle.intersects(textBlock.getPdfMinX(), textBlock.getPdfMinY(), textBlock.getWidth(), textBlock.getHeight()));
+    }
+
+}

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

@@ -152,7 +152,7 @@ public class PdfVisualisationUtility {
 
 
     @SneakyThrows
-    public static void drawLine2DList(PDDocument pdDocument, int pageNumber, List<Line2D> line2DS, Options options) {
+    public static void drawLine2DList(PDDocument pdDocument, int pageNumber, List<? extends Line2D> line2DS, Options options) {
 
         var pdPage = pdDocument.getPage(pageNumber - 1);
         var contentStream = new PDPageContentStream(pdDocument, pdPage, PDPageContentStream.AppendMode.APPEND, true);

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

@@ -19,9 +19,10 @@ public final class PositionUtils {
 
         double threshold = textBlock.getMostPopularWordHeight() * 3;
 
-        if (textBlock.getPdfMinX() + threshold > btf.getTopLeft().getX() && textBlock.getPdfMaxX() - threshold < btf.getTopLeft()
-                .getX() + btf.getWidth() && textBlock.getPdfMinY() + threshold > btf.getTopLeft().getY() && textBlock.getPdfMaxY() - threshold < btf.getTopLeft()
-                .getY() + btf.getHeight()) {
+        if (textBlock.getPdfMinX() + threshold > btf.getTopLeft().getX()
+                && textBlock.getPdfMaxX() - threshold < btf.getTopLeft().getX() + btf.getWidth()
+                && textBlock.getPdfMinY() + threshold > btf.getTopLeft().getY()
+                && textBlock.getPdfMaxY() - threshold < btf.getTopLeft().getY() + btf.getHeight()) {
             return true;
         } else {
             return false;

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

@@ -3,7 +3,6 @@ package com.knecon.fforesight.service.layoutparser.processor.utils;
 import java.util.List;
 import java.util.stream.Collectors;
 
-import org.apache.pdfbox.util.QuickSort;
 
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPageBlock;
 import com.knecon.fforesight.service.layoutparser.processor.model.text.TextPositionSequence;

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

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

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

@@ -1,52 +1,53 @@
 import org.springframework.boot.gradle.tasks.bundling.BootBuildImage
 
 plugins {
+    application
     id("com.knecon.fforesight.java-conventions")
-    id("org.springframework.boot") version "3.1.2"
-    id("io.spring.dependency-management") version "1.1.0"
-    id("org.sonarqube") version "4.2.1.3168"
-    id("io.freefair.lombok") version "8.1.0"
-    //id("org.graalvm.buildtools.native") version "0.9.23"
-}
-
-dependencies {
-    implementation(project(":layoutparser-service-processor"))
-
-    implementation("com.iqser.red.commons:storage-commons:2.1.0")
-    implementation("com.knecon.fforesight:tenant-commons:0.10.0")
-
-    implementation("org.springframework.boot:spring-boot-starter-actuator:3.1.2")
-    implementation("com.amazonaws:aws-java-sdk-s3:1.12.514")
-
-
-    // for integration testing only
-    testImplementation(project(":layoutparser-service-internal-api"))
-
-    testImplementation("org.springframework.boot:spring-boot-starter-amqp:3.0.6")
-    testImplementation("com.iqser.red.service:persistence-service-shared-api-v1:2.36.0")
-    testImplementation("com.iqser.red.commons:jackson-commons:1.0.0")
-    testImplementation("com.fasterxml.jackson.module:jackson-module-afterburner:2.15.0-rc2")
-    testImplementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:2.15.0-rc2")
-    testImplementation("org.apache.pdfbox:pdfbox:3.0.0-alpha2")
-    testImplementation("org.apache.pdfbox:pdfbox-tools:3.0.0-alpha2")
-    testImplementation("org.springframework.boot:spring-boot-starter-test:3.0.6")
-    testImplementation("org.apache.logging.log4j:log4j-slf4j-impl:2.19.0")
+    id("org.springframework.boot") version "3.1.3"
+    id("io.spring.dependency-management") version "1.1.3"
+    id("org.sonarqube") version "4.3.0.3225"
+    id("io.freefair.lombok") version "8.2.2"
+//    id("org.graalvm.buildtools.native") version "0.9.23"
 }
 
 description = "layoutparser-service-server"
 
-java {
-    withJavadocJar()
+val jacksonVersion = "2.15.2"
+val pdfBoxVersion = "3.0.0"
+
+dependencies {
+    implementation(project(":layoutparser-service-processor"))
+    implementation(project(":layoutparser-service-internal-api"))
+
+    implementation("com.iqser.red.commons:storage-commons:2.45.0")
+    implementation("com.knecon.fforesight:tenant-commons:0.14.0")
+    implementation("org.springframework.boot:spring-boot-starter-actuator:3.1.3")
+    implementation("com.amazonaws:aws-java-sdk-s3:1.12.536")
+    implementation("org.springframework.boot:spring-boot-starter-amqp:3.1.3")
+    implementation("org.springframework.cloud:spring-cloud-starter-openfeign:4.0.4")
+
+    // for integration testing only
+    testImplementation(project(":layoutparser-service-internal-api"))
+
+    testImplementation("org.springframework.boot:spring-boot-starter-amqp:3.1.3")
+    testImplementation("com.iqser.red.service:persistence-service-shared-api-v1:2.144.0")
+    testImplementation("com.iqser.red.commons:jackson-commons:1.3.0")
+    testImplementation("com.fasterxml.jackson.module:jackson-module-afterburner:${jacksonVersion}")
+    testImplementation("com.fasterxml.jackson.datatype:jackson-datatype-jsr310:${jacksonVersion}")
+    testImplementation("org.apache.pdfbox:pdfbox:${pdfBoxVersion}")
+    testImplementation("org.apache.pdfbox:pdfbox-tools:${pdfBoxVersion}")
+    testImplementation("org.springframework.boot:spring-boot-starter-test:3.1.3")
 }
 
 /*
-TODO: implement native build using GRAALVM, currently the problem is, logback needs runtime configuration using reflections or something
-kinda weird behavior, since the buildArgs below seem to have no effect at all
-https://stackoverflow.com/questions/72770461/graalvm-native-image-can-not-compile-logback-dependencies
+//TODO: implement native build using GRAALVM, currently the problem is, logback needs runtime configuration using reflections or something
+//kinda weird behavior, since the buildArgs below seem to have no effect at all
+//https://stackoverflow.com/questions/72770461/graalvm-native-image-can-not-compile-logback-dependencies
 
+We traced the issue to storage-commons and there to azure blob storage, this is today an open bug with azure spring
+https://github.com/Azure/azure-sdk-for-java/issues/36001
 
-// AOT seems to be the name of the generated classes for native images
-// They are added as a SourceSet, and therefore checkstyle and pmd try to run on it
+//AOT is springs ahead of time compilation, it is a code analysis which generates new code, therefore checkstyle and pmd can be disabled on them.
 tasks.named<Checkstyle>("checkstyleAot") {
     enabled = false
 }
@@ -59,29 +60,20 @@ tasks.named<Pmd>("pmdAot") {
 tasks.named<Pmd>("pmdAotTest") {
     enabled = false
 }
+*/
 
-graalvmNative {
-    binaries {
-        all {
-            resources.autodetect()
-            buildArgs.add("--verbose")
-            buildArgs.add("--add-opens=java.base/java.nio=ALL-UNNAMED")
-            buildArgs.add("--add-opens=java.base/jdk.internal.misc=ALL-UNNAMED")
-            buildArgs.add("--add-opens=java.base/jdk.internal.ref=ALL-UNNAMED")
-            buildArgs.add("--trace-class-initialization=ch.qos.logback.classic.Logger")
-            buildArgs.add("--trace-object-instantiation=ch.qos.logback.core.AsyncAppenderBase\$Worker")
-            buildArgs.add("--initialize-at-build-time=org.slf4j.LoggerFactory,ch.qos.logback")
-            buildArgs.add("--initialize-at-run-time=io.netty")
-        }
-
+sonarqube {
+    properties {
+        providers.gradleProperty("sonarToken").getOrNull()?.let { property("sonar.login", it) }
+        property("sonar.host.url", "https://sonarqube.knecon.com")
     }
 }
 
-nativeBuild {
-    buildArgs("-H:ReflectionConfigurationFiles=../../../src/main/resources/reflection-config.json")
-}
-*/
 tasks.named<BootBuildImage>("bootBuildImage") {
+
+    environment.put("BPE_DELIM_JAVA_TOOL_OPTIONS", " ")
+    environment.put("BPE_APPEND_JAVA_TOOL_OPTIONS", "-Dfile.encoding=UTF-8")
+
     imageName.set("nexus.knecon.com:5001/ff/${project.name}:${project.version}")
     if (project.hasProperty("buildbootDockerHostNetwork")) {
         network.set("host")

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

@@ -10,7 +10,7 @@ import org.springframework.context.annotation.Import;
 import com.amazonaws.services.s3.model.metrics.MetricsConfiguration;
 import com.iqser.red.storage.commons.StorageAutoConfiguration;
 import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingServiceProcessorConfiguration;
-import com.knecon.fforesight.service.layoutparser.processor.queue.MessagingConfiguration;
+import com.knecon.fforesight.service.layoutparser.server.queue.MessagingConfiguration;
 import com.knecon.fforesight.tenantcommons.MultiTenancyAutoConfiguration;
 
 @ImportAutoConfiguration({MultiTenancyAutoConfiguration.class})

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

@@ -1,7 +1,6 @@
-package com.knecon.fforesight.service.layoutparser.processor.queue;
+package com.knecon.fforesight.service.layoutparser.server.queue;
 
-import static com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingQueueNames.LAYOUT_PARSING_FINISHED_EVENT_QUEUE;
-import static com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingQueueNames.LAYOUT_PARSING_REQUEST_QUEUE;
+import java.util.Arrays;
 
 import org.springframework.amqp.AmqpRejectAndDontRequeueException;
 import org.springframework.amqp.core.Message;
@@ -12,7 +11,9 @@ import org.springframework.stereotype.Service;
 
 import com.fasterxml.jackson.databind.ObjectMapper;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingFinishedEvent;
+import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingQueueNames;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingRequest;
+import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
 import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingPipeline;
 
 import lombok.RequiredArgsConstructor;
@@ -30,12 +31,15 @@ public class MessageHandler {
 
 
     @RabbitHandler
-    @RabbitListener(queues = LAYOUT_PARSING_REQUEST_QUEUE)
+    @RabbitListener(queues = LayoutParsingQueueNames.LAYOUT_PARSING_REQUEST_QUEUE)
     @SneakyThrows
     public void receiveLayoutParsingRequest(Message message) {
 
         LayoutParsingRequest layoutParsingRequest = objectMapper.readValue(message.getBody(), LayoutParsingRequest.class);
 
+        if (layoutParsingRequest.layoutParsingType().equals(LayoutParsingType.TAAS) && layoutParsingRequest.researchDocumentStorageId() == null) {
+            throw new IllegalArgumentException("ResearchDocumentDataStorageId is null!");
+        }
         log.info("Layout parsing request received {}", layoutParsingRequest.identifier());
         if (message.getMessageProperties().isRedelivered()) {
             throw new AmqpRejectAndDontRequeueException(String.format("Error during last layout parsing of request with identifier: %s, do not retry.",
@@ -43,13 +47,13 @@ public class MessageHandler {
         }
         LayoutParsingFinishedEvent layoutParsingFinishedEvent = layoutParsingPipeline.parseLayoutAndSaveFilesToStorage(layoutParsingRequest);
         sendLayoutParsingFinishedEvent(layoutParsingFinishedEvent);
-        log.info("Layout parsing finished {} in {} ms", layoutParsingRequest.identifier(), layoutParsingFinishedEvent.duration());
     }
 
 
     public void sendLayoutParsingFinishedEvent(LayoutParsingFinishedEvent layoutParsingFinishedEvent) {
 
-        rabbitTemplate.convertAndSend(LAYOUT_PARSING_FINISHED_EVENT_QUEUE, layoutParsingFinishedEvent);
+        Arrays.stream(layoutParsingFinishedEvent.message().split("\n")).forEach(log::info);
+        rabbitTemplate.convertAndSend(LayoutParsingQueueNames.LAYOUT_PARSING_FINISHED_EVENT_QUEUE, layoutParsingFinishedEvent);
     }
 
 }

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

@@ -0,0 +1,38 @@
+package com.knecon.fforesight.service.layoutparser.server.queue;
+
+import org.springframework.amqp.core.Queue;
+import org.springframework.amqp.core.QueueBuilder;
+import org.springframework.context.annotation.Bean;
+import org.springframework.context.annotation.Configuration;
+
+import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingQueueNames;
+
+import lombok.RequiredArgsConstructor;
+
+@Configuration
+@RequiredArgsConstructor
+public class MessagingConfiguration {
+
+    @Bean
+    public Queue layoutparsingRequestQueue() {
+
+        return QueueBuilder.durable(LayoutParsingQueueNames.LAYOUT_PARSING_REQUEST_QUEUE)//
+                .withArgument("x-dead-letter-exchange", "").withArgument("x-dead-letter-routing-key", LayoutParsingQueueNames.LAYOUT_PARSING_DLQ).build();
+    }
+
+
+    @Bean
+    public Queue layoutparsingResponseQueue() {
+
+        return QueueBuilder.durable(LayoutParsingQueueNames.LAYOUT_PARSING_FINISHED_EVENT_QUEUE)//
+                .withArgument("x-dead-letter-exchange", "").withArgument("x-dead-letter-routing-key", LayoutParsingQueueNames.LAYOUT_PARSING_DLQ).build();
+    }
+
+
+    @Bean
+    public Queue layoutparsingDLQ() {
+
+        return QueueBuilder.durable(LayoutParsingQueueNames.LAYOUT_PARSING_DLQ).build();
+    }
+
+}

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

@@ -2,6 +2,7 @@ info:
   description: Layout Parser Service Processor
 
 tenant-user-management-service.url: "http://tenant-user-management-service:8080/internal"
+
 fforesight.tenants.remote: true
 
 server:

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

@@ -2,9 +2,9 @@ package com.knecon.fforesight.service.layoutparser.server;
 
 import org.junit.jupiter.api.Test;
 
-import com.knecon.fforesight.service.layoutparser.server.utils.BaseTest;
+import com.knecon.fforesight.service.layoutparser.server.utils.AbstractTest;
 
-class ApplicationTests extends BaseTest {
+class ApplicationTests extends AbstractTest {
 
     @Test
     void contextLoads() {

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

@@ -1,11 +1,11 @@
 package com.knecon.fforesight.service.layoutparser.server;
 
+import static org.assertj.core.api.AssertionsForClassTypes.assertThat;
+
 import java.awt.Color;
 import java.io.File;
-import java.io.FileInputStream;
 import java.io.FileOutputStream;
 import java.io.IOException;
-import java.io.InputStream;
 import java.nio.file.Files;
 import java.nio.file.Path;
 import java.nio.file.Paths;
@@ -17,22 +17,24 @@ import org.apache.pdfbox.pdmodel.PDDocument;
 import org.junit.jupiter.api.Disabled;
 import org.junit.jupiter.api.Test;
 import org.springframework.beans.factory.annotation.Autowired;
+import org.springframework.core.io.ClassPathResource;
 
 import com.fasterxml.jackson.databind.ObjectMapper;
 import com.knecon.fforesight.service.layoutparser.internal.api.data.taas.ResearchDocumentData;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
 import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingPipeline;
-import com.knecon.fforesight.service.layoutparser.processor.python_api.model.image.ImageServiceResponse;
-import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.textblock.TextBlock;
+import com.knecon.fforesight.service.layoutparser.processor.python_api.model.image.ImageServiceResponse;
+import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
+import com.knecon.fforesight.service.layoutparser.processor.services.factory.DocumentGraphFactory;
 import com.knecon.fforesight.service.layoutparser.processor.services.mapper.TaasDocumentDataMapper;
-import com.knecon.fforesight.service.layoutparser.server.utils.BaseTest;
+import com.knecon.fforesight.service.layoutparser.server.utils.AbstractTest;
 import com.knecon.fforesight.service.layoutparser.server.utils.visualizations.PdfDraw;
 
 import lombok.SneakyThrows;
 
-public class BdrJsonBuildTest extends BaseTest {
+public class BdrJsonBuildTest extends AbstractTest {
 
     @Autowired
     private ObjectMapper objectMapper;
@@ -42,16 +44,29 @@ public class BdrJsonBuildTest extends BaseTest {
 
 
     @SneakyThrows
-    protected Document buildGraph(File filename) {
+    protected Document buildGraph(File file) {
 
-        try (InputStream inputStream = new FileInputStream(filename)) {
-            try (PDDocument pdDocument = Loader.loadPDF(inputStream)) {
-                return layoutParsingPipeline.parseLayoutWithTimer(LayoutParsingType.TAAS, pdDocument, new ImageServiceResponse(), new TableServiceResponse());
-            }
+        try (PDDocument pdDocument = Loader.loadPDF(file)) {
+            return DocumentGraphFactory.buildDocumentGraph(layoutParsingPipeline.parseLayout(LayoutParsingType.TAAS,
+                    pdDocument,
+                    new ImageServiceResponse(),
+                    new TableServiceResponse()));
         }
     }
 
 
+    @Test
+    @SneakyThrows
+    public void testBDRFile199865() {
+
+        File file = new ClassPathResource("files/bdr/Drucksache_19_9865.pdf").getFile();
+        Document document = buildGraph(file);
+        ResearchDocumentData researchDocumentData = TaasDocumentDataMapper.fromDocument(document);
+        assertThat(researchDocumentData).isNotNull();
+
+    }
+
+
     @Test
     @Disabled
     public void writeBDRDocumentData() throws IOException {
@@ -95,10 +110,7 @@ public class BdrJsonBuildTest extends BaseTest {
 
     private static void visualizeSemanticNodes(File file, File resultingFileName, Document document, TextBlock textBlock) throws IOException {
 
-        try (var fileStream = new FileInputStream(file);//
-             PDDocument pdDocument = Loader.loadPDF(fileStream);//
-             var outputStream = new FileOutputStream(resultingFileName)//
-        ) {
+        try (PDDocument pdDocument = Loader.loadPDF(file); var outputStream = new FileOutputStream(resultingFileName)) {
             PdfDraw.drawDocumentGraph(pdDocument, document);
             PdfDraw.drawTextBlock(pdDocument, textBlock, PdfDraw.Options.builder().stroke(true).strokeWidth(0.1f).strokeColor(Color.YELLOW).build());
             pdDocument.save(outputStream);

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

@@ -30,10 +30,11 @@ import com.iqser.red.storage.commons.StorageAutoConfiguration;
 import com.iqser.red.storage.commons.service.StorageService;
 import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
 import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingPipeline;
-import com.knecon.fforesight.service.layoutparser.processor.python_api.model.image.ImageServiceResponse;
-import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.Document;
 import com.knecon.fforesight.service.layoutparser.processor.model.graph.nodes.SemanticNode;
+import com.knecon.fforesight.service.layoutparser.processor.python_api.model.image.ImageServiceResponse;
+import com.knecon.fforesight.service.layoutparser.processor.python_api.model.table.TableServiceResponse;
+import com.knecon.fforesight.service.layoutparser.processor.services.factory.DocumentGraphFactory;
 import com.knecon.fforesight.tenantcommons.TenantsClient;
 
 import lombok.AllArgsConstructor;
@@ -94,10 +95,10 @@ public class HeadlinesGoldStandardIntegrationTest {
         goldStandardLog.getRedactionLogEntry().removeIf(r -> !r.isRedacted() || r.getChanges().get(r.getChanges().size() - 1).getType().equals(ChangeType.REMOVED));
         goldStandardLog.getRedactionLogEntry().forEach(e -> goldStandardHeadlines.add(new Headline(e.getPositions().get(0).getPage(), e.getValue())));
 
-        Document documentGraph = layoutParsingPipeline.parseLayout(LayoutParsingType.REDACT_MANAGER,
-                Loader.loadPDF(pdfFileResource.getInputStream()),
+        Document documentGraph = DocumentGraphFactory.buildDocumentGraph(layoutParsingPipeline.parseLayout(LayoutParsingType.REDACT_MANAGER,
+                Loader.loadPDF(pdfFileResource.getFile()),
                 new ImageServiceResponse(),
-                new TableServiceResponse());
+                new TableServiceResponse()));
 
         var foundHeadlines = documentGraph.streamAllSubNodes()
                 .map(SemanticNode::getHeadline)

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

@@ -0,0 +1,34 @@
+package com.knecon.fforesight.service.layoutparser.server;
+
+import java.util.Arrays;
+
+import org.junit.jupiter.api.Test;
+import org.springframework.beans.factory.annotation.Autowired;
+
+import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingFinishedEvent;
+import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingRequest;
+import com.knecon.fforesight.service.layoutparser.internal.api.queue.LayoutParsingType;
+import com.knecon.fforesight.service.layoutparser.processor.LayoutParsingPipeline;
+import com.knecon.fforesight.service.layoutparser.server.utils.AbstractTest;
+
+import lombok.SneakyThrows;
+import lombok.extern.slf4j.Slf4j;
+
+@Slf4j
+public class LayoutparserEnd2EndTest extends AbstractTest {
+
+    @Autowired
+    private LayoutParsingPipeline layoutParsingPipeline;
+
+
+    @Test
+    @SneakyThrows
+    public void testLayoutParserEndToEnd() {
+
+        prepareStorage("files/bdr/btd_19_053_1905391.pdf");
+        LayoutParsingRequest layoutParsingRequest = buildDefaultLayoutParsingRequest(LayoutParsingType.TAAS);
+        LayoutParsingFinishedEvent finishedEvent = layoutParsingPipeline.parseLayoutAndSaveFilesToStorage(layoutParsingRequest);
+        Arrays.stream(finishedEvent.message().split("\n")).forEach(log::info);
+    }
+
+}