Phone number

Imagine you are validating a phone number, and we know the number will be in the format 123-XXX-XXXX. Here are a couple of alternatives.

const phoneNumber = '123-760-9001'

We can check if the phone number is a string with length 12.

expect(phoneNumber, '12 chars')
  .to.be.a('string')
  .and.to.have.length(12)

Unfortunately, the assertion is way too forgiving.

expect('WRONG PHONE!', 'also has 12 chars')
  .to.be.a('string')
  .and.to.have.length(12)

Maybe we need to use a regular expression. We know the phone number can only contain digits and - characters. Let's use the match(regex) assertion. Plus we can use ^ and $ characters to make sure we match the entire string.

expect(phoneNumber, 'digits and -').to.match(/^[\d-]{12}$/)

Ok, this looks better, but it still allows for some funny values.

expect('------------', 'just -').to.match(/^[\d-]{12}$/)
expect('0-0-0-0-0-0-', 'caterpillar').to.match(/^[\d-]{12}$/)
expect('123456789012', 'wrong number of digits').to.match(
  /[\d-]{12}/,
)

My advice write the strictest regular expression possible enforcing the type and position of each character class. In our case it is the following regular expression:

const phoneRegex = /^\d\d\d-\d\d\d-\d\d\d\d$/
expect(phoneNumber, 'phone #').to.match(phoneRegex)

If you want, you can write it differently:

const phoneRegex = /^\d{3}-\d{3}-\d{4}$/

All previous funny inputs fail our strict regular expression.

expect('WRONG PHONE!', 'also has 12 chars').to.not.match(
  phoneRegex,
)
expect('------------', 'just -').to.not.match(phoneRegex)
expect('0-0-0-0-0-0-', 'caterpillar').to.not.match(phoneRegex)
expect('123456789012', 'wrong number of digits').to.not.match(
  phoneRegex,
)

# Phone number

Phone number