The problem is that writing a parser for a file format as complex as PDF in an unsafe language is just asking for trouble. There were many exploits already that

The problem is that writing a parser for a file format as complex as PDF in an unsafe language is just asking for trouble. There were many exploits already that involved a specially crafted PDF file to gain RCE, and there would probably be more, unless they do rewrite it. At this point we should just accept that it's next to impossible to write a 100% safe parser in an unsafe language.