When using regexes to "parse" html it is often useful to use a negative character class instead of .*? as all the RE engine has to do is grab stuff up to the next tag. This will of course choke if closing tags are muddled.

m/<tag>([^<]*)</tag>/
[download]