Fri Dec 27 13:14:23 EST 2013

Splitting regexp -> view?

Many real world tables contain "fields within fields" that make more
sense when they are separated out.  Otherwise queries involve many
regular expressions.

What is bettter: to leave the somewhat inconsistent relations burried
inside the fields, or to separate them out into separate fields.

Maybe the real question is: is it possible to JOIN on a regular
expression relation?

And if the answer is yes, is it too inefficient?

I see no reason why it would not be possible.  Joining should work on
any predicate.  If so then it might not be necessary to do a lot of
splitting (= premature optimization?).

Is it possible to store regexes in tables?