CODEWITHSUNDEEP
regex
Trevor
Trevor

Reputation: 2457

Combine regex statements together

I'm trying to combine these regular expressions into one. Anybody know if this is possible?

Basically I'm trying to check if a string has a lowercase letter, an uppercase letter and a number.

(.*[a-z]+.*)
(.*[A-Z]+.*)
(.*[0-9]+.*)

So these strings that would all pass

aA1
a1A
Aa1
A1a
1aA
1Aa

Examples that would return false

aa1
AA1
a1a
AAa
a1
1a
1AA
1A
etc..

Upvotes: 2

Views: 91

Answers (2)

Schwern
Schwern

Reputation: 164769

Yes, but it's nasty. (Examples are in Perl)

my $re = qr{(?:
    ([a-z]+) |
    ([A-Z]+) |
    ([0-9]+)
)+}x;

All of $1, $2, and $3 must be defined (true won't cut it because 0 is false).

$string =~ $re;
print "Match" if defined $1 && defined $2 && defined $3;

You're better off for everyone's sake keeping it at three simple regexes and writing a subroutine.

sub is_valid_string {
    my $str = shift;

    return 0 unless $str =~ /[a-z]/;
    return 0 unless $str =~ /[A-Z]/;
    return 0 unless $str =~ /[0-9]/;

    return 1;
}

Upvotes: 1

Sergey Kalinichenko
Sergey Kalinichenko

Reputation: 726539

A very simple approach is to enumerate all six combinations:

[a-z][A-Z][0-9]|[a-z][0-9][A-Z]|[A-Z][a-z][0-9]|[A-Z][0-9][a-z]|[0-9][a-z][A-Z]|[0-9][A-Z][a-z]

Essentially, you've got characters of three types - Upper, Lower, and Digit. The six combinations are

LUD|LDU|ULD|UDL|DLU|DUL

As you can see, this gets complex pretty quickly. If you are dong it for password checking, doing character class counting in the host language is a better choice.

Upvotes: 1

Related Questions