How might we improve credit card filtering?

[stevepolitodesign]

This regex only matches 16-digit credit cards but will miss valid 15-digit American Express cards and 14-digit Diners Club cards. Consider using a more comprehensive pattern like /\b[3456]\d{13,15}\b/ or separate patterns for different card types.

  # Matches Visa (16), MasterCard (16), Discover (16), Amex (15), Diners Club (14)
  CREDIT_CARD_REGEX = /
    \b(
      3[47]\d{13} |                # American Express (15 digits)
      3(?:0[0-5]|[68]\d)\d{11} |  # Diners Club (14 digits)
      4\d{12}(?:\d{3})? |         # Visa (13 or 16 digits)
      5[1-5]\d{14} |              # MasterCard (16 digits)
      6(?:011|5\d{2})\d{12}       # Discover (16 digits)
    )
  \b
  /x
  # Matches same as above, but with delimiters (space, +, or -)
  CREDIT_CARD_REGEX_DELIMITERS = /
    \b(
      3[47][\s+-]?\d{4}[\s+-]?\d{6}[\s+-]?\d{5} |                # Amex
      3(?:0[0-5]|[68]\d)[\s+-]?\d{2}[\s+-]?\d{6}[\s+-]?\d{4} |  # Diners Club
      4[\s+-]?\d{3}[\s+-]?\d{4}[\s+-]?\d{4}[\s+-]?\d{1,4} |     # Visa
      5[1-5][\s+-]?\d{2}[\s+-]?\d{4}[\s+-]?\d{4}[\s+-]?\d{4} |  # MasterCard
      6(?:011|5\d{2})[\s+-]?\d{4}[\s+-]?\d{4}[\s+-]?\d{4}       # Discover
    )
  \b
  /x

Originally posted by @Copilot in #9 (comment)

[stevepolitodesign]

I think the best solution (for now) is either

• Train your existing model.
• Supply your own regex with a custom filter Support custom filters #31