Here the regex VALID_EMAIL_REGEX is a constant, indicated in Ruby by a name starting with a capital letter. The code

ensures that only email addresses that match the pattern will be considered valid. (Because it starts with a capital letter, VALID_EMAIL_REGEX is a Ruby constant, so its value can’t change.)

                So, where does the pattern come from? Regular expressions consist of a terse (some would say unreadable) language for matching text patterns; learning to construct regexes is an art, and to get you started I’ve broken VALID_EMAIL_REGEX into bite-sized pieces (Table 6.1).10 To really learn about regular expressions, though, I consider the amazing Rubular regular expression editor (Figure 6.4) to be simply essential.11 The Rubular website has a beautiful interactive interface for making regular expressions, along with a handy regex quick reference. I encourage you to study Table 6.1 with a browser window open to Rubular—no amount of reading about regular expressions can replace a couple of hours playing with Rubular. (Note: If you use the regex from Listing 6.17 in Rubular, you should leave off the \A and \z characters.)

By the way, there actually exists a full regex for matching email addresses according to the official standard, but it’s really not worth the trouble. The one in Listing 6.17 is
fine, maybe even better than the official one.12

                  The tests should all be passing now. (In fact, the tests for valid email addresses should have been passing all along; since regexes are notoriously error-prone, the valid
email tests are there mainly as a sanity check on VALID_EMAIL_REGEX.) This means that there’s only one constraint left: enforcing the email addresses to be unique.