Bug: non-native PDFs don't get text extracted when processed

[aexshafii]

solution: process it as an image when that is the case