[PDFBOX-4878] Call to DictionaryEncoding.getEncoding () throws NullPointerException for some PDF's

• Type: Bug
• Status: Closed
• Priority: Critical
• Resolution: Fixed
• Affects Version/s: 2.0.20
• Fix Version/s: 2.0.21, 3.0.0 PDFBox
• Component/s: PDModel
• Assignee: Tilman Hausherr
• Reporter: Thomas Gawehn
• Created: 12/Jun/20 13:58
• Updated: 21/Aug/20 04:54
• Resolved: 10/Jul/20 07:01

Description

Wenn iterating all fonts in a PDF a call to PDSimpleFont.getEncoding().getEncodingName () throws a NullPointerException inside the library code for the attached sample-PDF.

The situation can be reproduced, when running the following loop for the attached PDF (test.pdf):

 

 

        PDDocument document = PDDocument.load (file);
        int pages = document.getNumberOfPages ();
        for (int i = 0; i < pages; ++i)
        {
            PDPage page = document.getPage (i);
            PDResources resources = page.getResources ();
            if (resources == null)
                continue;
            Iterator<COSName> iter = resources.getFontNames ().iterator ();
            while (iter.hasNext ())
            {
                COSName cos = iter.next ();
                try
                {
                    PDFont font = resources.getFont (cos);
                    System.out.println (font.getName ());
                    if (font instanceof PDSimpleFont)
                    {
                        PDSimpleFont simpleFont = (PDSimpleFont) font;
                        Encoding encoding = simpleFont.getEncoding ();
                        if (encoding != null)
                        {
                            try
                            {
                                System.out.println ("* encoding="+encoding.getEncodingName ());
                            }
                            catch (Exception e)
                            {
                                e.printStackTrace ();
                            }
                        }
                    }
                }
                catch (IOException e)
                {
                    e.printStackTrace ();
                }
            }
        }
        document.close ();