garbled CJK Unified Ideographs Extension text test and fix (#987)

* add CJK Unified Ideographs Extension B, C, D test * Fixed garbled text * fix unused import
LibrePDF · Nov 15, 2023 · 5e46ab0 · 5e46ab0
1 parent 4f3ad5b
commit 5e46ab0
Show file tree

Hide file tree

Showing 2 changed files with 79 additions and 4 deletions.
diff --git a/openpdf/src/main/java/com/lowagie/text/pdf/FontDetails.java b/openpdf/src/main/java/com/lowagie/text/pdf/FontDetails.java
@@ -239,7 +239,7 @@ byte[] convertToBytes(String text, TextRenderingOptions options) {
     private byte[] convertToBytesWithGlyphs(String text) throws UnsupportedEncodingException {
         int len = text.length();
         int[] metrics = null;
-        char[] glyph = new char[len];
+        int[] glyph = new int[len];
         int i = 0;
         for (int k = 0; k < len; ++k) {
             int val;
@@ -257,10 +257,19 @@ private byte[] convertToBytesWithGlyphs(String text) throws UnsupportedEncodingE
             Integer gl = m0;
             if (!longTag.containsKey(gl))
                 longTag.put(gl, new int[]{m0, metrics[1], val});
-            glyph[i++] = (char)m0;
+            glyph[i++] = m0;
         }
-        String s = new String(glyph, 0, i);
-        return s.getBytes(CJKFont.CJK_ENCODING);
+        return getCJKEncodingBytes(glyph, i);
+    }
+
+    private byte[] getCJKEncodingBytes(int[] glyph, int size) {
+        byte[] result = new byte[size * 2];
+        for (int i = 0; i < size; i++) {
+            int g = glyph[i];
+            result[i * 2] = (byte)(g >> 8);
+            result[i * 2 + 1] = (byte)(g & 0xFF);
+        }
+        return result;
     }
 
     byte[] convertToBytes(GlyphVector glyphVector) {

diff --git a/openpdf/src/test/java/com/lowagie/text/pdf/PdfDocumentCJKExtensionTest.java b/openpdf/src/test/java/com/lowagie/text/pdf/PdfDocumentCJKExtensionTest.java
@@ -0,0 +1,66 @@
+package com.lowagie.text.pdf;
+
+import java.io.FileOutputStream;
+import java.io.IOException;
+
+import org.junit.jupiter.api.Test;
+
+import com.lowagie.text.Chunk;
+import com.lowagie.text.Document;
+import com.lowagie.text.Font;
+import com.lowagie.text.FontFactory;
+
+class PdfDocumentCJKExtensionTest {
+	@Test
+	void generateDocumentsWithCJKExtension() throws IOException {
+		String fontName = "TakaoMjMincho";
+
+		// TakaoMjMincho Version 003.01.01
+		// Please download and place it below. 
+		// https://launchpad.net/takao-fonts
+		// https://launchpad.net/takao-fonts/trunk/15.03/+download/TakaoMjFonts_00301.01.zip
+		String fontPath = "src/test/resources/fonts/TakaoMjFonts/TakaoMjMincho.ttf";
+
+		// register font
+		FontFactory.register(fontPath, fontName);
+
+		Document document = new Document();
+
+		// FOP off 
+		document.setGlyphSubstitutionEnabled(false);
+
+		PdfWriter.getInstance(document, new FileOutputStream("target/" + PdfDocumentCJKExtensionTest.class.getSimpleName() + ".pdf"));
+
+		try {
+			Font font = FontFactory.getFont(fontName, "Identity-H", false, 10, 0, null);
+
+			document.open();
+			// CJK Unified Ideographs Extension B
+			// https://en.wikipedia.org/wiki/CJK_Unified_Ideographs_Extension_B
+
+			// U+20000
+			String cjkB_OK = "𠀀";
+			// U+2A000 
+			String cjkB_NG = "𪀀";
+
+
+			// CJK Unified Ideographs Extension C
+			// https://en.wikipedia.org/wiki/CJK_Unified_Ideographs_Extension_C
+
+			// U+2A746
+			String cjkC_NG = "𪝆";
+
+
+			// CJK Unified Ideographs Extension D
+			// https://en.wikipedia.org/wiki/CJK_Unified_Ideographs_Extension_D
+
+			// U+2B746
+			String cjkD_NG = "𫝆";
+
+			document.add(new Chunk(cjkB_OK + " " + cjkB_NG + " " + cjkC_NG + " " + cjkD_NG, font));
+
+		} finally {
+			document.close();
+		}
+	}
+}