in reply to Parsing web pages (sort of)

In case it was not clear from previous posts, your regex was fine, but you matched and then did nothing. You needed to print the match. So your regex:
m/\bDE\d+/g
would match and stop. What they did (in addition to being more specific with the number of digits) is to group your match in parentheses:
m/\b(DE\d+)/g
And then the match is located in $1. (Which they print)