Searchable PDFs created from scanned Arabic document images > look good, the test is pretty accurate, but the text layered on the > image in the PDF has a fatal flaw: no (or insufficient) space between > words. As a result searching and copying are useless. Is there any > chance this will be looked into, or is their a fix that can be applied > to the PDFs?
I assume you are trying to copy and paste the text directly from the overlay?
This issue seems specific to Arabic OCR, so thanks for the hint. We will be looking into it.