[RELEASE] iText pdfOCR 3.0.2

itext · Jan 30, 2024 · 78c4b90 · 78c4b90
2 parents da196c6 + 151cd11
commit 78c4b90
Show file tree

Hide file tree

Showing 136 changed files with 1,204 additions and 183 deletions.
diff --git a/pdfocr-api/pom.xml b/pdfocr-api/pom.xml
@@ -5,7 +5,7 @@
   <parent>
     <groupId>com.itextpdf</groupId>
     <artifactId>pdfocr-root</artifactId>
-    <version>3.0.1</version>
+    <version>3.0.2</version>
   </parent>
 
   <artifactId>pdfocr-api</artifactId>

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/AbstractPdfOcrEventHelper.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/AbstractPdfOcrEventHelper.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/IImageRotationHandler.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/IImageRotationHandler.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/IOcrEngine.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/IOcrEngine.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/IOcrProcessProperties.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/IOcrProcessProperties.java
@@ -0,0 +1,29 @@
+/*
+    This file is part of the iText (R) project.
+    Copyright (c) 1998-2024 Apryse Group NV
+    Authors: Apryse Software.
+
+    This program is offered under a commercial and under the AGPL license.
+    For commercial licensing, contact us at https://itextpdf.com/sales.  For AGPL licensing, see below.
+
+    AGPL licensing:
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU Affero General Public License as published by
+    the Free Software Foundation, either version 3 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU Affero General Public License for more details.
+
+    You should have received a copy of the GNU Affero General Public License
+    along with this program.  If not, see <https://www.gnu.org/licenses/>.
+ */
+package com.itextpdf.pdfocr;
+
+/**
+ * OCR properties passed to the OCR engine as part of {@link OcrProcessContext}.
+ */
+public interface IOcrProcessProperties {
+}
diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/IProductAware.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/IProductAware.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/OcrEngineProperties.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/OcrEngineProperties.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/OcrPdfCreator.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/OcrPdfCreator.java
diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/OcrPdfCreatorEventHelper.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/OcrPdfCreatorEventHelper.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/OcrPdfCreatorProperties.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/OcrPdfCreatorProperties.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.
@@ -99,6 +99,11 @@ public class OcrPdfCreatorProperties {
 
     private IMetaInfo metaInfo;
 
+    /**
+     * Indicates whether the created pdf is tagged or not.
+     */
+    private boolean tagged = false;
+
     /**
      * Creates a new {@link OcrPdfCreatorProperties} instance.
      */
@@ -369,6 +374,26 @@ public OcrPdfCreatorProperties setImageRotationHandler(
         return this;
     }
 
+    /**
+     * Defines whether pdf document should be tagged or not.
+     *
+     * @param tagged {@code true} if the result pdf is expected to be tagged, {@code false} otherwise.
+     * @return this {@link OcrPdfCreatorProperties} instance.
+     */
+    public OcrPdfCreatorProperties setTagged(boolean tagged) {
+        this.tagged = tagged;
+        return this;
+    }
+
+    /**
+     * Retrieve information on whether pdf document should be tagged or not.
+     *
+     * @return {@code true} if the result pdf is expected to be tagged, {@code false} otherwise.
+     */
+    public boolean isTagged() {
+        return tagged;
+    }
+
     /**
      * Set meta info for this {@link OcrPdfCreatorProperties}.
      *

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/OcrProcessContext.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/OcrProcessContext.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.
@@ -28,6 +28,8 @@ This file is part of the iText (R) project.
 public class OcrProcessContext {
     private AbstractPdfOcrEventHelper ocrEventHelper;
 
+    private IOcrProcessProperties ocrProcessProperties;
+
     /**
      * Creates an instance of ocr process context
      *
@@ -54,4 +56,22 @@ public AbstractPdfOcrEventHelper getOcrEventHelper() {
     public void setOcrEventHelper(AbstractPdfOcrEventHelper eventHelper) {
         this.ocrEventHelper = eventHelper;
     }
+
+    /**
+     * Set extra OCR process properties.
+     *
+     * @param ocrProcessProperties extra OCR process properties.
+     */
+    void setOcrProcessProperties(IOcrProcessProperties ocrProcessProperties) {
+        this.ocrProcessProperties = ocrProcessProperties;
+    }
+
+    /**
+     * Get extra OCR process properties.
+     *
+     * @return extra OCR process properties.
+     */
+    public IOcrProcessProperties getOcrProcessProperties() {
+        return ocrProcessProperties;
+    }
 }
diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/PdfCreatorUtil.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/PdfCreatorUtil.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.
@@ -96,6 +96,7 @@ static float calculateFontSize(final Document document, final String line,
 
         try {
             Paragraph paragraph = new Paragraph(line);
+            paragraph.setMargin(0);
             paragraph.setWidth(bbox.getWidth());
             paragraph.setFontFamily(fontFamily);
 

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/PdfOcrFontProvider.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/PdfOcrFontProvider.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/PdfOcrMetaInfoContainer.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/PdfOcrMetaInfoContainer.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/ScaleMode.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/ScaleMode.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/TextInfo.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/TextInfo.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.
@@ -23,11 +23,7 @@ This file is part of the iText (R) project.
 package com.itextpdf.pdfocr;
 
 import com.itextpdf.kernel.geom.Rectangle;
-
-import java.util.ArrayList;
-import java.util.Collections;
-import java.util.List;
-
+import com.itextpdf.pdfocr.structuretree.LogicalStructureTreeItem;
 
 /**
  * This class describes how recognized text is positioned on the image
@@ -45,6 +41,11 @@ public class TextInfo {
      */
     private Rectangle bboxRect;
 
+    /**
+     * If LogicalStructureTreeItem is set, then {@link TextInfo}s are expected to be in logical order.
+     */
+    private LogicalStructureTreeItem logicalStructureTreeItem;
+
     /**
      * Creates a new {@link TextInfo} instance.
      */
@@ -107,4 +108,27 @@ public Rectangle getBboxRect() {
     public void setBboxRect(final Rectangle bbox) {
         this.bboxRect = new Rectangle(bbox);
     }
+
+    /**
+     * Retrieves structure tree item for the text item.
+     *
+     * @return structure tree item.
+     */
+    public LogicalStructureTreeItem getLogicalStructureTreeItem() {
+        return logicalStructureTreeItem;
+    }
+
+    /**
+     * Sets logical structure tree parent item for the text info. It allows to organize text chunks
+     * into logical hierarchy, e.g. specify document paragraphs, tables, etc.
+     * <p>
+     *
+     * If LogicalStructureTreeItem is set, then the list of {@link TextInfo}s in {@link IOcrEngine#doImageOcr}
+     * return value is expected to be in logical order.
+     *
+     * @param logicalStructureTreeItem structure tree item.
+     */
+    public void setLogicalStructureTreeItem(LogicalStructureTreeItem logicalStructureTreeItem) {
+        this.logicalStructureTreeItem = logicalStructureTreeItem;
+    }
 }
diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/exceptions/PdfOcrException.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/exceptions/PdfOcrException.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/exceptions/PdfOcrExceptionMessageConstant.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/exceptions/PdfOcrExceptionMessageConstant.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.
@@ -33,6 +33,7 @@ public class PdfOcrExceptionMessageConstant {
     public static final String CANNOT_CREATE_PDF_DOCUMENT = "Cannot create PDF document: {0}";
     public static final String STATISTICS_EVENT_TYPE_CANT_BE_NULL = "Statistics event type can't be null";
     public static final String STATISTICS_EVENT_TYPE_IS_NOT_DETECTED = "Statistics event type is not detected.";
+    public static final String TAGGING_IS_NOT_SUPPORTED = "Tagging is not supported by the OCR engine.";
 
     private PdfOcrExceptionMessageConstant() {
         //Private constructor will prevent the instantiation of this class directly

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/exceptions/PdfOcrInputException.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/exceptions/PdfOcrInputException.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/logs/PdfOcrLogMessageConstant.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/logs/PdfOcrLogMessageConstant.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/statistics/PdfOcrOutputType.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/statistics/PdfOcrOutputType.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/...pi/src/main/java/com/itextpdf/pdfocr/statistics/PdfOcrOutputTypeStatisticsAggregator.java b/...pi/src/main/java/com/itextpdf/pdfocr/statistics/PdfOcrOutputTypeStatisticsAggregator.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/statistics/PdfOcrOutputTypeStatisticsEvent.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/statistics/PdfOcrOutputTypeStatisticsEvent.java
@@ -1,6 +1,6 @@
 /*
     This file is part of the iText (R) project.
-    Copyright (c) 1998-2023 Apryse Group NV
+    Copyright (c) 1998-2024 Apryse Group NV
     Authors: Apryse Software.
 
     This program is offered under a commercial and under the AGPL license.

diff --git a/pdfocr-api/src/main/java/com/itextpdf/pdfocr/structuretree/ArtifactItem.java b/pdfocr-api/src/main/java/com/itextpdf/pdfocr/structuretree/ArtifactItem.java
@@ -0,0 +1,44 @@
+/*
+    This file is part of the iText (R) project.
+    Copyright (c) 1998-2024 Apryse Group NV
+    Authors: Apryse Software.
+
+    This program is offered under a commercial and under the AGPL license.
+    For commercial licensing, contact us at https://itextpdf.com/sales.  For AGPL licensing, see below.
+
+    AGPL licensing:
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU Affero General Public License as published by
+    the Free Software Foundation, either version 3 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU Affero General Public License for more details.
+
+    You should have received a copy of the GNU Affero General Public License
+    along with this program.  If not, see <https://www.gnu.org/licenses/>.
+ */
+package com.itextpdf.pdfocr.structuretree;
+
+/**
+ * This class represents artifact structure tree item. Attaching such item to the text info means that
+ * the text will be marked as artifact.
+ */
+public final class ArtifactItem extends LogicalStructureTreeItem {
+    private final static ArtifactItem ARTIFACT_INSTANCE = new ArtifactItem();
+
+    private ArtifactItem() {
+        super();
+    }
+
+    /**
+     * Retrieve an instance of {@link ArtifactItem}.
+     *
+     * @return an instance of {@link ArtifactItem}.
+     */
+    public static ArtifactItem getInstance() {
+        return ARTIFACT_INSTANCE;
+    }
+}