asp.net-corepdf-generationpdfsharpmigradocpdfsharpcore

How to use unicode characters in MigraDoc in Linux

In latest Migradoc 6.0.0-preview-3, some unicode characters in PDF file appear as not accented in Linux. To to reproduce, run in Debian in ASP.NET MVC controller

  static ReportTemplateBase()
  {
    GlobalFontSettings.FontResolver = new NewFontResolver();
  }

  var textFrame = Section.AddTextFrame();
  var font = new Font("Arial", 12);
  var par = textFrame.AddParagraph();
  par.Format.Font = font;
  par.AddFormattedText("õäöüÕÄÖÜ", font);

If running in Debian Linux, äöü are replaced with unaccented characters. Output is

õaouÕAOU

If running from Visual Studio IDE in Windows, output is correct:

õäöüÕÄÖÜ

How to get correct characters in Debian Linux?

Looked into NewFontResolver source code in https://github.com/empira/PDFsharp/blob/91feca7521ad073cbfd132d4f957525824fbac17/src/foundation/src/shared/src/PdfSharp.Snippets/Font/fontresolving/NewFontResolver.cs#L186

It searches font files in Linux directories.

I copied .ttf files to /usr/share/fonts/fontsforpdfcreation directory and it looks like this solved the issue.

Is this OK?

Solution

  • Create your own FontResolver implementing IFontResolver and include TTF files that have all the characters you need. Arial will do as it works under Windows.