From 5e602e9b24342aff36e3d08fe8e1b28e0214d0f9 Mon Sep 17 00:00:00 2001
From: James Welch <james.welch@bdi.ox.ac.uk>
Date: Mon, 1 Jul 2024 12:49:49 +0100
Subject: [PATCH] Update test and use PdfBox to provide additional validation

---
 dependencies.gradle                           |  1 +
 .../dita/processor/DitaProcessorSpec.groovy   | 22 +++++++++++++++++--
 2 files changed, 21 insertions(+), 2 deletions(-)

diff --git a/dependencies.gradle b/dependencies.gradle
index 8966f61..f7bd2a0 100644
--- a/dependencies.gradle
+++ b/dependencies.gradle
@@ -94,6 +94,7 @@ dependencies {
     //
     runtimeOnly group: 'ant-contrib', name: 'ant-contrib', version: '0.6'
 
+    testImplementation 'org.apache.pdfbox:pdfbox:2.0.27'
 
     // For the html helper functions
     api group: 'net.sf.jtidy', name: 'jtidy', version: 'r938'
diff --git a/src/test/groovy/uk/ac/ox/softeng/dita/processor/DitaProcessorSpec.groovy b/src/test/groovy/uk/ac/ox/softeng/dita/processor/DitaProcessorSpec.groovy
index 181e970..8e94017 100644
--- a/src/test/groovy/uk/ac/ox/softeng/dita/processor/DitaProcessorSpec.groovy
+++ b/src/test/groovy/uk/ac/ox/softeng/dita/processor/DitaProcessorSpec.groovy
@@ -21,11 +21,16 @@ import uk.ac.ox.softeng.maurodatamapper.dita.DitaProject
 import uk.ac.ox.softeng.maurodatamapper.dita.elements.langref.base.DitaMap
 import uk.ac.ox.softeng.maurodatamapper.dita.elements.langref.base.Topic
 import uk.ac.ox.softeng.maurodatamapper.dita.processor.DitaProcessor
+
+import groovy.util.logging.Slf4j
+import org.apache.pdfbox.pdmodel.PDDocument
+import org.apache.pdfbox.text.PDFTextStripper
 import spock.lang.Specification
 
 import java.nio.file.Files
 import java.nio.file.Paths
 
+@Slf4j
 class DitaProcessorSpec extends Specification{
 
     DitaProcessor ditaProcessor
@@ -61,11 +66,24 @@ class DitaProcessorSpec extends Specification{
         }
 
         byte[] fileContents = ditaProcessor.generatePdf(ditaProject)
+        log.debug("File contents size: ${fileContents.size()}")
+        then:
+        fileContents.size() > 7700 // The number of bytes of the generated pdf file
+        fileContents.size() < 7800
+
+
+
+        when: // Use PdfBox to validate the PDF
+
         Files.write(Paths.get('build/tmp/pdftest.pdf'), fileContents)
+        PDDocument pdDocument = PDDocument.load(new File('build/tmp/pdftest.pdf'))
+        PDFTextStripper pdfStripper = new PDFTextStripper()
+        String text = pdfStripper.getText(pdDocument)
+        log.debug(text)
 
-        System.err.println("File contents size: ${fileContents.size()}")
         then:
-        fileContents.size() == 7771 // The number of bytes of the generated pdf file
+        text.contains("My first topic")
+        text.contains("Hello, World!")
 
     }