Upgrade apache pdfbox to version 3

digipost · Jan 5, 2024 · b9cda43 · b9cda43
1 parent 2a875a2
commit b9cda43
Show file tree

Hide file tree

Showing 4 changed files with 33 additions and 10 deletions.
diff --git a/NOTICE b/NOTICE
@@ -1,17 +1,25 @@
 Digipost Printability Validator
 
-Copyright 2023 Posten Norge AS. All Rights Reserved.
+Copyright 2024 Posten Norge AS. All Rights Reserved.
 
 This product includes software developed by Posten Norge AS. - https://www.posten.no/
 Licensed under Apache 2 - http://www.apache.org/licenses/LICENSE-2.0.html
 
 
 This software includes third party software subject to the following licenses:
 
-  Apache FontBox under Apache License, Version 2.0
-  Apache PDFBox under Apache License, Version 2.0
+  Apache FontBox under Apache-2.0
+  Apache PDFBox under Apache-2.0
+  Apache PDFBox io under Apache-2.0
   Digipost Printability Validator under The Apache Software License, Version 2.0
   JCL 1.2 implemented over SLF4J under Apache License, Version 2.0
+  JUnit Jupiter (Aggregator) under Eclipse Public License v2.0
+  JUnit Jupiter Engine under Eclipse Public License v2.0
+  JUnit Jupiter Params under Eclipse Public License v2.0
+  JUnit Platform Commons under Eclipse Public License v2.0
+  JUnit Platform Engine API under Eclipse Public License v2.0
+  org.apiguardian:apiguardian-api under The Apache License, Version 2.0
+  org.opentest4j:opentest4j under The Apache License, Version 2.0
   SLF4J API Module under MIT License
 
 
diff --git a/pom.xml b/pom.xml
@@ -33,7 +33,7 @@
     <properties>
         <maven.compiler.source>1.8</maven.compiler.source>
         <maven.compiler.target>1.8</maven.compiler.target>
-        <slf4j-api.version>2.0.9</slf4j-api.version>
+        <slf4j-api.version>2.0.10</slf4j-api.version>
     </properties>
 
     <dependencies>
@@ -57,7 +57,18 @@
         <dependency>
             <groupId>org.apache.pdfbox</groupId>
             <artifactId>pdfbox</artifactId>
-            <version>2.0.27</version>
+            <version>3.0.1</version>
+            <exclusions>
+                <exclusion>
+                    <groupId>commons-logging</groupId>
+                    <artifactId>commons-logging</artifactId>
+                </exclusion>
+            </exclusions>
+        </dependency>
+        <dependency>
+            <groupId>org.apache.pdfbox</groupId>
+            <artifactId>pdfbox-io</artifactId>
+            <version>3.0.1</version>
             <exclusions>
                 <exclusion>
                     <groupId>commons-logging</groupId>
@@ -219,7 +230,8 @@
                                         <include>org.slf4j</include>
                                         <include>org.apache.commons:commons-lang3</include>
                                         <include>org.hamcrest</include>
-                                        <include>org.junit.jupiter:junit-jupiter-api</include>
+                                        <include>org.junit.jupiter:*</include>
+                                        <include>org.junit.platform:*</include>
                                         <include>org.opentest4j:opentest4j</include>
                                         <include>org.apiguardian:apiguardian-api</include>
                                         <include>org.junit.platform:junit-platform-commons</include>

diff --git a/src/NOTICE.template b/src/NOTICE.template
@@ -1,6 +1,6 @@
 Digipost Printability Validator
 
-Copyright 2023 Posten Norge AS. All Rights Reserved.
+Copyright 2024 Posten Norge AS. All Rights Reserved.
 
 This product includes software developed by Posten Norge AS. - https://www.posten.no/
 Licensed under Apache 2 - http://www.apache.org/licenses/LICENSE-2.0.html

diff --git a/src/main/java/no/digipost/print/validate/PdfValidator.java b/src/main/java/no/digipost/print/validate/PdfValidator.java
@@ -16,6 +16,9 @@
 package no.digipost.print.validate;
 
 import no.digipost.print.validate.PdfValidationSettings.Bleed;
+import org.apache.pdfbox.Loader;
+import org.apache.pdfbox.io.RandomAccessInputStream;
+import org.apache.pdfbox.io.RandomAccessReadBuffer;
 import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.pdmodel.common.PDRectangle;
@@ -85,7 +88,7 @@ public PdfValidationResult validate(Path pdfFile, PdfValidationSettings printVal
     private PdfValidationResult validateForPrint(InputStream pdfStream, PdfValidationSettings printValidationSettings) {
         int numberOfPages = -1;
         List<PdfValidationError> errors;
-        try (PDDocument pdDoc = PDDocument.load(pdfStream)) {
+        try (PDDocument pdDoc = Loader.loadPDF(new RandomAccessReadBuffer(pdfStream))) {
             numberOfPages = pdDoc.getNumberOfPages();
             errors = validateDocumentForPrint(pdDoc, printValidationSettings);
         } catch (InvalidPasswordException invalidPassword) {
@@ -195,7 +198,7 @@ private void validerSideantall(int numberOfPages, int maxPages, final List<PdfVa
         }
         if (numberOfPages == 0) {
             errors.add(PdfValidationError.DOCUMENT_HAS_NO_PAGES);
-            LOG.debug("The PDF document does not contain any pages. The file may be corrupt.", numberOfPages);
+            LOG.debug("The PDF document does not contain any pages. The file may be corrupt.");
         }
     }
 
@@ -243,7 +246,7 @@ private boolean hasTextInArea(PDPage pdPage, Rectangle2D area) throws IOExceptio
         stripper.addRegion("marginArea", area);
         stripper.extractRegions(pdPage);
         String text = stripper.getTextForRegion("marginArea");
-        if (text != null && text.trim().length() > 0) {
+        if (text != null && !text.trim().isEmpty()) {
             hasTextInArea = true;
         }
         return hasTextInArea;