Regular expression character class classes vs. shorthands

[Zhris]

Hi,

I have 2 alternative regexes to check for a valid UK Postcode. I know that they probably do not check accurately, however I believe the 2 regexes are identical but do not produce the same output.

[blue]Regex 1 (classes) (works as expected):[/blue]
if($_=~s/^([A-Za-z]{1,3}[0-9]{1}|[A-Za-z]{1,2}[0-9]{2})([0-9]{1}[A-Za-z]{2})$/$1 $2/){print "$_ > Passed<br>";}

[blue]Regex 2 (shorthands) (does not work as expected):[/blue]
if($_=~s/^(\w{1,3}\d{1}|\w{1,2}\d{2})(\d{1}\w{2})$/$1 $2/){print "$_ > Passed<br>";}

Regex 2 will pass postcodes such as 'AA1 11A' and 'AA1 1A1' which should fail.

What have I done wrong / why do they pass?

Thanks,

Chris

 

[KevinADC]

its because \w includes the digits 0-9.

\w = [a-zA-Z0-9_]

------------------------------------------
- Kevin, perl coder unexceptional!

 

[Zhris]

Ahh, thanks Kevin. I misread my quick reference. Therefore there is no shorthand and I have to use the character class [A-Za-z]?

Chris

 

[KevinADC]

Right, there is no short hand for A-Za-z so you have to use the character class.

------------------------------------------
- Kevin, perl coder unexceptional!