G'day kosie99,

The short answer is to write it both ways and then Benchmark.

I have, in the past, found leaving the matching to Perl to be faster; however, I have not had to do this with MySQL so I can't really provide specific advice. Furthermore, that was quite a long time in the past - things may have changed.

For matching simple substrings, without any fancy pattern matching requirements, consider index instead of using a regular expression.