Angular reactive forms pattern validator adding $ to regex and breaking validation

Question

I'm trying to validate a password with the following regex:
^(?=.*[A-Z])(?=.*[a-z])(?=.*[0-9])(?=.{8,}).
Note that there is no terminating $ char, because that would prevent valid passwords from being accepted. (I'm not sure why the input field doesn't terminate the string).
However, Angular's Validators.pattern adds the end of string char. And therefore my valid passwords fail.
How do I prevent the pattern validator from adding the $?
I suppose I could roll my own password validator, but surely there is a better solution...?

EDIT: Passwords that should succeed:

Test1234
tEst1234
tesT1234
1234Test
1234tesT
t#St1234

Passwords that should fail:

TEST1234
test1234
tEst123
Test123
testtest
12345678

Validator declaration:

this.password = new FormControl('', [Validators.required, Validators.pattern('^(?=.*[A-Z])(?=.*[a-z])(?=.*[0-9])(?=.{8,})')]);
this.userForm.addControl('Password', this.password);

Please give us one or two examples of what passwords you tried, and also how you declared your validator. — user4676340
– user4676340, Commented Aug 8, 2018 at 8:07

Wiktor Stribiżew · Accepted Answer · 2020-02-19 15:30:18Z

30

You may add a .* at the end, or even revamp the pattern a bit to convert one lookahead into a consuming pattern:

Validators.pattern('(?=.*[A-Z])(?=.*[a-z])(?=.*[0-9]).{8,}')

Or, better, when using a regex for password validation, follow the principle of contrast:

Validators.pattern('(?=[^A-Z]*[A-Z])(?=[^a-z]*[a-z])(?=[^0-9]*[0-9]).{8,}')

Angular will add ^ and $ on both ends of the regex pattern, so the pattern will look like

^(?=[^A-Z]*[A-Z])(?=[^a-z]*[a-z])(?=[^0-9]*[0-9]).{8,}$

Note it won't add the anchors automatically if you use a regex literal, add them manually:

Validators.pattern(/^(?=[^A-Z]*[A-Z])(?=[^a-z]*[a-z])(?=\D*\d).{8,}$/)

With a regex literal, you may use single backslashes in regex escapes (/\d/ to match a digit vs. "\\d" in a string literal).

See the regex demo

Details

^ - start of string
(?=[^A-Z]*[A-Z]) - at least 1 uppercase ASCII letter
(?=[^a-z]*[a-z]) - at least 1 lowercase ASCII letter
(?=[^0-9]*[0-9]) - at least 1 ASCII digit
.{8,} - any 8 or more chars (other than line break chars)
$ - end of string.

edited Feb 19, 2020 at 15:30

answered Aug 8, 2018 at 8:08

Wiktor Stribiżew

631k41 gold badges501 silver badges629 bronze badges

Sign up to request clarification or add additional context in comments.

11 Comments

AvailableName Over a year ago

Changing the length lookahead to a consuming pattern solved it. Thanks.

Jignesh Mistry Over a year ago

Can i assign this dynamically i mean (?=[^A-Z]*[A-Z]) this code validates for 1 upper case letter what if i want 2 uppercase letters ?

Wiktor Stribiżew Over a year ago

@JigneshMistry Replace (?=[^A-Z]*[A-Z]) with (?=(?:[^A-Z]*[A-Z]){2})

Jignesh Mistry Over a year ago

Thanks buddy it worked :) Just one more trouble can u send the regex for 2 Numeric & 2 special characters ?

Wiktor Stribiżew Over a year ago

@JigneshMistry They are formed the same way as above.

|

Collectives™ on Stack Overflow

Angular reactive forms pattern validator adding $ to regex and breaking validation

1 Answer 1

11 Comments

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

11 Comments

Linked

Related