in reply to RE: RE: RE: Regular Expression Optimizer
in thread Regular Expression Optimizer
I will concede that there may be a case where two equally optimal REs are available. This is entirely plausible, and I sometimes fall prey to a binary train of thought. Thus, while I've been reading replies to all of this, I've thought more as I go along, and since I did want to spur dialogue, I've let my thoughts wander down the ways everyone else is thinking to see how approachable this could be. Which of course leads me to your final two questions:
Obviously, this whole discussion of language theory has mathematical rules under which it operates. I think everyone here is aware of what a regular expression optimizer runs up against simply by the fact that no one was able to say "Oh, you want to go look at x." So, my question was simply if heuristic manipulations would be allowed to the input RE. Basically, what if during implementation, a few "leap of faith" type heuristics were involved, but they also performed remarkably well. Does that affect the reception of my REO machine into the community since it doesn't always act as expected mathematically? ALL HAIL BRAK!!!
|
|---|