Here is the code that will be in release 2.0.14 in the examples subproject:
/**
* Output a text without knowing which font is the right one. One use case is a worldwide
* address list. Only LTR languages are supported, RTL (e.g. Hebrew, Arabic) are not
* supported so they would appear in the wrong direction.
* Complex scripts (Thai, Arabic, some Indian languages) are also not supported, any output
* will look weird. There is an (unfinished) effort here:
* https://issues.apache.org/jira/browse/PDFBOX-4189
*
* @author Tilman Hausherr
*/
public class EmbeddedMultipleFonts
{
public static void main(String[] args) throws IOException
{
try (PDDocument document = new PDDocument())
{
PDPage page = new PDPage(PDRectangle.A4);
document.addPage(page);
PDFont font1 = PDType1Font.HELVETICA; // always have a simple font as first one
TrueTypeCollection ttc2 = new TrueTypeCollection(new File("c:/windows/fonts/batang.ttc"));
PDType0Font font2 = PDType0Font.load(document, ttc2.getFontByName("Batang"), true); // Korean
TrueTypeCollection ttc3 = new TrueTypeCollection(new File("c:/windows/fonts/mingliu.ttc"));
PDType0Font font3 = PDType0Font.load(document, ttc3.getFontByName("MingLiU"), true); // Chinese
PDType0Font font4 = PDType0Font.load(document, new File("c:/windows/fonts/mangal.ttf")); // Indian
PDType0Font font5 = PDType0Font.load(document, new File("c:/windows/fonts/ArialUni.ttf")); // Fallback
try (PDPageContentStream cs = new PDPageContentStream(document, page))
{
cs.beginText();
List<PDFont> fonts = new ArrayList<>();
fonts.add(font1);
fonts.add(font2);
fonts.add(font3);
fonts.add(font4);
fonts.add(font5);
cs.newLineAtOffset(20, 700);
showTextMultiple(cs, "abc 한국 中国 भारत 日本 abc", fonts, 20);
cs.endText();
}
document.save("example.pdf");
}
}
static void showTextMultiple(PDPageContentStream cs, String text, List<PDFont> fonts, float size)
throws IOException
{
try
{
// first try all at once
fonts.get(0).encode(text);
cs.setFont(fonts.get(0), size);
cs.showText(text);
return;
}
catch (IllegalArgumentException ex)
{
// do nothing
}
// now try separately
int i = 0;
while (i < text.length())
{
boolean found = false;
for (PDFont font : fonts)
{
try
{
String s = text.substring(i, i + 1);
font.encode(s);
// it works! Try more with this font
int j = i + 1;
for (; j < text.length(); ++j)
{
String s2 = text.substring(j, j + 1);
if (isWinAnsiEncoding(s2.codePointAt(0)) && font != fonts.get(0))
{
// Without this segment, the example would have a flaw:
// This code tries to keep the current font, so
// the second "abc" would appear in a different font
// than the first one, which would be weird.
// This segment assumes that the first font has WinAnsiEncoding.
// (all static PDType1Font Times / Helvetica / Courier fonts)
break;
}
try
{
font.encode(s2);
}
catch (IllegalArgumentException ex)
{
// it's over
break;
}
}
s = text.substring(i, j);
cs.setFont(font, size);
cs.showText(s);
i = j;
found = true;
break;
}
catch (IllegalArgumentException ex)
{
// didn't work, will try next font
}
}
if (!found)
{
throw new IllegalArgumentException("Could not show '" + text.substring(i, i + 1) +
"' with the fonts provided");
}
}
}
static boolean isWinAnsiEncoding(int unicode)
{
String name = GlyphList.getAdobeGlyphList().codePointToName(unicode);
if (".notdef".equals(name))
{
return false;
}
return WinAnsiEncoding.INSTANCE.contains(name);
}
}
Alternatives to arialuni can be found here:
https://en.wikipedia.org/wiki/Open-source_Unicode_typefaces
