I was trying to follow up on the issues in our system and found that some PDF files threw ERRORs. These PDFs are produced by a publishing system and that system seems to add fonts when you change to them and add them even though they are never used. Or only space is used. Then they add this font with an empty glyf table. This results in that errors are thrown on files that are fine.
Line 310 in TTFParser removes empty glyf tables.
// skip tables with zero length
if (table.getLength() == 0)
Line 215 of TTFParser throws exception when glyf table is missing.
if (font.getGlyph() == null)