Good idea. Maybe making the replacements short, space-separated words would have helped even more, by countering any "helpful corrections" the OCR might have been making based on English words (e.g., it would not surprise me if some OCR software would "correct" "B0B" to "BOB").