Hello!
I'm importing quite large text document (legacy-db-exported stuff) into database. each row consist of about 20 columns, parsed using PCRE, but NEEDS these ungreedy sort of things ... and one line takes up to 50 (!!!) second to parse, resulting in multi-day show. While it's needed (synchronization) to do it every 3-7 days, it's not that good. I ofcourse can hack it (like differential/update etc.), but I would really love to hear any suggestion on getting better performance of the PCRE itself.
thanks in advance,
regards,
Jakub Marecek