It's hard to give a really robust solution here without knowing what forms the data can take. First, do you always have Firstname Lastname, or do some players have middle names? Or suppose bell hooks or e.e. cummings starts to play baseball, you'll have to deal with case. But here's how you might do it. I assume you're getting the lines out as an array, and the second element is the player's name.

# split line into @data, or whatever
$data[1] =~ s/^([A-Z])\w*( \w+)$/$1.$2/g;
[download]

What that does is grab the first capital letter of $data[1] and remembers it in $1; then it matches a whole bunch of other letters (which are going to get thrown away), followed by a space and the last string of letters in the string (which it remembers as $2).

It then substitutes that whole thing for the first thing it remembered, followed by the period, then the second thing it remembered.

See perldoc perlre for more ways to have fun with regular expressions.

Note: this is only one way to do it =)

HTH

$fullname =~ /^(.)\S*\s(.+)/;
$fullname = "$1. $2";
[download]

I believe the technical aspects of the question are adequately addressed above--I'd go for the capital letters character class, personally.

That said, you're going to have trouble with this, one way or another--bell hooks doesn't play in the National League, but J.T. Snow does, as do Bobby Jones and Bobby Jones, not to mention two each of C. Jones and T. Jones (okay, one of those plays in the AL...). Martínez will give you similar issues, and Rodríguez is a mild pain as well (as is Smith, while we're at it).

Sure you can't deal with the long names?

Of course, I don't think any of those pairs play for the same team these days, so you're pretty safe if you stick to box scores. But I'd still be nervous doing it.