I'm trying to write a regular expression in perl to pull student usernames from input. Our campus usernames consist of the first five letters of our last name, followed by first initial, followed by middle initial (x if the person doesn't have one), followed a number representing the year they started. This number used to be single digit, but new ones are now 2. So as an example, my username is fitzpjm3.

Some things to consider here:
- some last names don't have more than 5 letters, so usernames will vary in length
- there is an optional 0 at the end of the username (like in robbiet04)

The expression I have been building on looks like
This searches, from beginning of string to end, for any number of letters, followed by a digit that may or may not have a leading 0. The only trouble I am having is that this matches pure digits. So I get a positive return on a username, but a digit will come back as true as well. How do I get it to only match when everything is true?

Well, I was able to resolve the issue by testing to see if $1 (the first part of the expression) exists. By doing it this way, if it finds a digit without letters before it, it won't recognize it as a username.

For example:
This seems a bit obtuse though. If there is a better alternative, please share.

The reason it is returning a match on pure digits has to do with the * quantifier on the first part. [A-Za-z]* will match zero or more letters, while [A-Za-z]+ will match one or more letters. Changing the * to a + should fix it. I find it somewhat rare to use *, because it always matches.

Also, beware of [A-z], you are including some non-letters in there (check out an ASCII chart to see what I mean).

Ah, yes, it worked. I totally overlooked the fact that * returns even on 0 matches. As for the A-z, thanks. I had done that on purpose thinking Z was followed by a. It was specifically because \w includes special characters that I chose to approach it that way. Been a while since I had to work with ASCII...