SPIKE: Normalize ISBNs for matching purposes

Apply these rules in the following matching scenarios (using the ISBN normalization utility we already built):

Incoming record type = MARC Bib, and Field = 020, and Subfield = a or z
Existing record type = MARC Bib, and Field = 020, and Subfield = a or z
Existing record type = Instance, and Field = Identifier, and Identifier type = ISBN or Invalid ISBN

Any 10-digit ISBN should be normalized to the 13-digit version before matching (note that 10 digit ISBNs can end in X, but 13-digit ISBNs are numeric-only)

• 047144250X normalizes to 9780471442509
  Any hyphens within the ISBN should be stripped
• 978-0-471-44250-9 normalizes to 9780471442509
  Any extra spaces before, internal to, or after the ISBN should be stripped
• 978 0 471 44250 9 normalizes to 9780471442509
  Any alpha-numeric qualifying data after the ISBN should be removed
• 9780471442509 (cloth : alk. paper) normalizes to 9780471442509

Only after all of the above, then try to match an incoming and existing ISBN. That means any combination of the above, e.g. an ISBN 10 with hyphens can be matched to an ISBN-13 without hyphens but with a qualifier

• 047144250X matches to 9780471442509
• 978-0-471-44250-9 matches to 9780471442509
• 047144250X (cloth : alk. paper) matches to 9780471442509
• 047144250X matches to 9780471442509 (vol. 1)

[and will need for add Existing record type = Order, and Field = Identifier, and Identifier type = ISBN, once we start importing orders]