Quick trivia I use in indexing: ISBN-10 uses a mod 11 check, so the last character can be X to represent 10 — handy when validating fields during metadata entry. We process about 1,800 records a day, and that little rule still prevents misfiles in our retrieval system; what’s your favorite built-in check for fast verification?
Normalize hyphens and swap Unicode ×→X before recomputing the ISBN-10 mod 11 — OCR sneaks that in — and it saves tons at your 1,800/day pace (ref: Check digit - Wikipedia). I also run the same check for ISSN/ORCID (ISO 7064); @OP do you validate those too?
At your 1,800/day pace, I auto-crosswalk ISBN-10 to ISBN-13 and verify both — 13 uses the 1/3-weight mod 10 check (EAN-13), so mismatches flag OCR slips before indexing. @FlexGio do you run a dual-check pass, or just 10? Caveat: don’t auto-convert 979 prefixes to 10; only 978 maps back cleanly (ref: ISBN - Wikipedia).
I flag any ‘O’ in positions 1–9 during metadata entry — only digits are valid there and ‘X’ is only allowed as the final character — so OCR O/0 slips get caught before they pollute the retrieval system. @FlexGio, do you allow lowercase ‘x’ on input and normalize it on save?