PDFBOX-6256: Add ActualText to GlyphLayoutBidiTest

pdfbox-layout-awt/src/test/java/org/apache/pdfbox/glyphlayout/awt/GlyphLayoutBidiTest.java

@@ -16,16 +16,27 @@
  */
 package org.apache.pdfbox.glyphlayout.awt;
 
+import org.apache.pdfbox.Loader;
+import org.apache.pdfbox.pdmodel.AbstractGlyphLayoutProcessor;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.pdmodel.PDPage;
+import org.apache.pdfbox.pdmodel.PDPageContentStream;
+import org.apache.pdfbox.pdmodel.font.PDType0Font;
+import org.apache.pdfbox.text.PDFTextStripper;
 import org.junit.jupiter.api.Test;
 
 import java.awt.FontFormatException;
+import java.io.BufferedWriter;
+import java.io.File;
+import java.io.FileOutputStream;
 import java.io.IOException;
+import java.io.OutputStream;
+import java.io.OutputStreamWriter;
+import java.io.Writer;
 import java.net.URISyntaxException;
+import java.nio.charset.StandardCharsets;
 
-import org.apache.pdfbox.pdmodel.PDDocument;
-import org.apache.pdfbox.pdmodel.PDPage;
-import org.apache.pdfbox.pdmodel.PDPageContentStream;
-import org.apache.pdfbox.pdmodel.font.PDType0Font;
+import static org.junit.jupiter.api.Assertions.assertEquals;
 
 /**
  * Examples for bidirectional text with GlyphLayoutProcessorAwt
@@ -73,37 +84,111 @@ private float showLine(PDPageContentStream cs, PDType0Font[] fonts, float fontSi
         return y;
     }
 
+    /**
+     * Test, no ActualText
+     *
+     * @throws IOException
+     * @throws FontFormatException
+     * @throws URISyntaxException
+     */
+    @Test
+    void testGlyphLayoutBidiNoActualText() throws IOException, FontFormatException, URISyntaxException
+    {
+        testGlyphLayoutBidi(false, "");
+    }
+
+    /**
+     * Test with ActualText
+     *
+     * @throws IOException
+     * @throws FontFormatException
+     * @throws URISyntaxException
+     */
     @Test
-    void testGlyphLayoutBidi() throws IOException, FontFormatException, URISyntaxException
+    void testGlyphLayoutBidiActualText() throws IOException, FontFormatException, URISyntaxException
     {
-        GlyphLayoutProcessorAwt glyphLayoutProcessorAwt = new GlyphLayoutProcessorAwt();
+        testGlyphLayoutBidi(true, "_ActualText");
+    }
+
+    /**
+     * Test
+     * @param useActualText
+     * @throws IOException
+     * @throws FontFormatException
+     * @throws URISyntaxException
+     */
+    void testGlyphLayoutBidi(boolean useActualText, String sActualText) throws IOException, FontFormatException, URISyntaxException
+    {
+        AbstractGlyphLayoutProcessor.GlyphLayoutProcessorOptions options = new AbstractGlyphLayoutProcessor.GlyphLayoutProcessorOptions();
+        if (useActualText)
+        {
+            options.useActualText();
+        }
+        GlyphLayoutProcessorAwt glyphLayoutProcessor = new GlyphLayoutProcessorAwt(options);
+
+        String outputBaseName = String.format("GlyphLayoutBidi%s", sActualText);
+        String outputPDFFilename = "target/" + outputBaseName + ".pdf";
+        String outputTextFilename = "target/" + outputBaseName + ".txt";
 
-        String outputName = "GlyphLayoutBidi.pdf";
-        String outputFilename = "target/" + outputName;
         String arabicPath = "/ttf/NotoSansArabic-Regular.ttf";
         String lgcPath = "/ttf/DejaVuSans.ttf";
 
         float fontSize = 12.0f;
 
         try (PDDocument doc = new PDDocument())
         {
-            PDType0Font arabicFont = createPdType0Font(glyphLayoutProcessorAwt, doc, arabicPath);
-            PDType0Font lgcFont = createPdType0Font(glyphLayoutProcessorAwt, doc, lgcPath);
+            PDType0Font arabicFont = createPdType0Font(glyphLayoutProcessor, doc, arabicPath);
+            PDType0Font lgcFont = createPdType0Font(glyphLayoutProcessor, doc, lgcPath);
 
             PDPage page = new PDPage();
             doc.addPage(page);
             try (PDPageContentStream cs = new PDPageContentStream(doc, page))
             {
-                cs.setGlyphLayoutProcessor(glyphLayoutProcessorAwt);
+                cs.setGlyphLayoutProcessor(glyphLayoutProcessor);
 
                 float x = page.getBBox().getLowerLeftX() + fontSize;
                 float y = page.getBBox().getUpperRightY() - fontSize;
 
                 y = showLine(cs, arabicFont, fontSize, x, y, TEXT1);
                 showLine(cs, new PDType0Font[]{ lgcFont, arabicFont, lgcFont }, fontSize, x, y, new String[]{ TEXT2, TEXT3, TEXT4 });
             }
-            doc.save(outputFilename);
+            doc.save(outputPDFFilename);
+        }
+
+        checkRenderIdent(outputBaseName + ".pdf");
+
+        // Extract text
+        try (PDDocument doc = Loader.loadPDF(new File(outputPDFFilename))) {
+            assertEquals(1, doc.getNumberOfPages());
+
+            PDFTextStripper stripper = new PDFTextStripper();
+            String s = stripper.getText(doc);
+            String sStripped = s.replace("\r", "").replaceAll(" +", " ")
+                    .replace(" \n", "\n")
+                    .strip();
+
+            String text =
+                    TEXT1 + "\n" + TEXT2 + TEXT3 + TEXT4;
+
+            if (useActualText) {
+                // Not correct as of 2026-09-27
+                // PDFTextStripper should take the text from ActualText as is and not reorder the text
+                assertEquals(text, sStripped, "Extracted text should equal the written text for " + outputPDFFilename);
+            } else {
+                // Extracted text is wrong
+                assertEquals(text, sStripped, "Extracted text should equal the written text for " + outputPDFFilename);
+            }
+
+            try (OutputStream os = new FileOutputStream(outputTextFilename)) {
+                os.write(0xEF);
+                os.write(0xBB);
+                os.write(0xBF);
+
+                try (Writer writer = new BufferedWriter(new OutputStreamWriter(os, StandardCharsets.UTF_8))) {
+                    // The output is not yet correct as of 27.9.2026
+                    writer.write(s);
+                }
+            }
         }
-        checkRenderIdent(outputName);
     }
 }

pdfbox-layout-awt/src/test/java/org/apache/pdfbox/glyphlayout/awt/GlyphLayoutDin91379FormTest.java

@@ -17,20 +17,31 @@
 package org.apache.pdfbox.glyphlayout.awt;
 
 import java.awt.FontFormatException;
+import java.io.BufferedWriter;
 import java.io.File;
+import java.io.FileOutputStream;
 import java.io.IOException;
 import java.io.InputStream;
+import java.io.OutputStream;
+import java.io.OutputStreamWriter;
+import java.io.Writer;
 import java.net.URISyntaxException;
+import java.nio.charset.StandardCharsets;
+
 import org.apache.pdfbox.Loader;
+import org.apache.pdfbox.pdmodel.AbstractGlyphLayoutProcessor;
 import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDDocumentCatalog;
 import org.apache.pdfbox.pdmodel.PDResources;
 import org.apache.pdfbox.pdmodel.font.PDType0Font;
 import org.apache.pdfbox.pdmodel.interactive.form.PDAcroForm;
 import org.apache.pdfbox.pdmodel.interactive.form.PDField;
 import org.apache.pdfbox.pdmodel.interactive.form.PDTextField;
+import org.apache.pdfbox.text.PDFTextStripper;
 import org.junit.jupiter.api.Test;
 
+import static org.junit.jupiter.api.Assertions.assertEquals;
+
 /*
  * Example of formatting for letters defined in:
  * DIN 91379: Characters and defined character sequences in Unicode for the electronic processing of names
@@ -84,15 +95,51 @@ class GlyphLayoutDin91379FormTest extends TestBase
                     + "ƒ ʰ ʳ ˆ ˜ ˢ ᵈ ᵗ ‘ ‚ “ ” „ † … ‰ ′ ″ ‹ › ⁰ ⁴ ⁵ ⁶ ⁷ ⁸ "
                     + "⁹ ⁿ ₀ ₁ ₂ ₃ ₄ ₅ ₆ ₇ ₈ ₉ ™ ∞ ≤ ≥ \n"
                     + "Additional non-letters (not included in DIN 91379): – — •�";
+    /**
+     * Test, no ActualText
+     *
+     * @throws IOException
+     * @throws FontFormatException
+     * @throws URISyntaxException
+     */
+    @Test
+    void testGlyphLayoutDin91379FormNoActualText() throws IOException, FontFormatException, URISyntaxException
+    {
+        testGlyphLayoutDin91379Form(false, "");
+    }
 
+    /**
+     * Test with ActualText
+     *
+     * @throws IOException
+     * @throws FontFormatException
+     * @throws URISyntaxException
+     */
     @Test
-    void testGlyphLayoutDin91379Form() throws IOException, FontFormatException, URISyntaxException
+    void testGlyphLayoutDin91379FormActualText() throws IOException, FontFormatException, URISyntaxException
     {
-        GlyphLayoutProcessorAwt glyphLayoutProcessor = new GlyphLayoutProcessorAwt();
+        testGlyphLayoutDin91379Form(true, "_ActualText");
+    }
 
-        String outputName = "GlyphLayoutDIN91379Form.pdf";
-        String outputFilename = "target/" + outputName;
+    /**
+     * Test
+     * @param useActualText
+     * @throws IOException
+     * @throws FontFormatException
+     * @throws URISyntaxException
+     */
+    void testGlyphLayoutDin91379Form(boolean useActualText, String sActualText) throws IOException, FontFormatException, URISyntaxException
+    {
+        AbstractGlyphLayoutProcessor.GlyphLayoutProcessorOptions options = new AbstractGlyphLayoutProcessor.GlyphLayoutProcessorOptions();
+        if (useActualText)
+        {
+            options.useActualText();
+        }
+        GlyphLayoutProcessorAwt glyphLayoutProcessor = new GlyphLayoutProcessorAwt(options);
 
+        String outputBaseName = String.format("GlyphLayoutDIN91379Form%s", sActualText);
+        String outputPDFFilename = "target/" + outputBaseName + ".pdf";
+        String outputTextFilename = "target/" + outputBaseName + ".txt";
         String fontSizeString = "12";
 
         InputStream fontStream = GlyphLayoutDin91379FormTest.class.getResourceAsStream("/ttf/Arimo-Regular.ttf");
@@ -125,8 +172,47 @@ void testGlyphLayoutDin91379Form() throws IOException, FontFormatException, URIS
             }
             acroForm.refreshAppearances();
             acroForm.flatten();
-            doc.save(outputFilename);
+            doc.save(outputPDFFilename);
+        }
+
+        checkRenderIdent(outputBaseName + ".pdf");
+
+        // Extract text
+        try (PDDocument doc = Loader.loadPDF(new File(outputPDFFilename))) {
+            assertEquals(1, doc.getNumberOfPages());
+
+            PDFTextStripper stripper = new PDFTextStripper();
+            String s = stripper.getText(doc);
+            String sStripped = s.replace('\n', ' ')
+                    .replaceAll(" +", " ")
+                    .replace(" ", "\n")
+                    .strip();
+
+            String expectedText = ("Test form for PDFBox glyph layout\n" + LATIN_CHARS_DIN_91379)
+                    .replace('\n', ' ')
+                    .replaceAll(" +", " ")
+                    .replace(" ", "\n")
+                    .strip();
+
+            try (OutputStream os = new FileOutputStream(outputTextFilename)) {
+                os.write(0xEF);
+                os.write(0xBB);
+                os.write(0xBF);
+
+                try (Writer writer = new BufferedWriter(new OutputStreamWriter(os, StandardCharsets.UTF_8))) {
+                    // The output without ActualText is not yet correct
+                    writer.write(s);
+                }
+            }
+
+            if (useActualText) {
+                // ok with ActualText
+                assertEquals(expectedText, sStripped, "Extracted text should equal the written text for " + outputPDFFilename);
+            } else {
+                // not ok without ActualText
+                assertEquals(expectedText, sStripped, "Extracted text should equal the written text for " + outputPDFFilename);
+            }
+
         }
-        checkRenderIdent(outputName);
     }
 }