[PDFBOX-1742] type1CFont font with null encoding

• Type: Bug
• Status: Closed
• Priority: Major
• Resolution: Fixed
• Affects Version/s: 1.8.3, 2.0.0
• Fix Version/s: 2.0.0
• Component/s: PDModel
• Assignee: Andreas Lehmkühler
• Reporter: Tilman Hausherr
• Created: 08/Oct/13 17:05
• Updated: 17/Mar/16 19:08
• Resolved: 19/Jan/14 11:12

Description

The file from PDFBOX-427 does not display the bullets (near "diff" in the right column)

Through debugging I found out:

The glyph code to be rendered is 183. That code doesn't exist in the font, however a code 116 exists, and it came from an "sid" code (CFFGlyph2D constructor). A mapping of the name ("bullet") through the nameToCode table isn't done because encoding is null.

PDSimpleFont.determineEncoding() sets a null encoding
PDType1Font.determineEncoding() sets a "good" encoding based on getAFM()
(both called from constructor)
getAFM() gets an encoding table that includes codetoname/nametocode 183 <=> bullet

Later, due to having a FontFile3, a new type1CFont is created with the external file, but with no encoding.

The following change in PDType1Font.java works for me (first line is the existing line):

if (type1CFont.getFontEncoding() == null && getFontEncoding() != null) //TH
{
    LOG.info("Encoding of Type1CFont '" + type1CFont.getBaseFont() + "' is null, trying Type1Font encoding");
    type1CFont.setFontEncoding(getFontEncoding());
    }

Another improvement is the fat dash (at the beginning of "--use-merge-history" at the bottom of the mid column.

Issue Links

• is related to: PDFBOX-427 ArrayIndexOutOfBoundsException in drawString
• is related to: PDFBOX-1844 [PATCH] Parser for Type 1 Fonts