Sorry for the late reply -- I thought I would get a notice of some sort if someone posted in my thread... and I just looked.
newt good point, although regex will still be slow against a properly indexed database (and what exactly would your regex be, say, to find two or more 'A's and an 'E' anywhere in any word? -- here is my run against 45000 words ... so there might be some speed/simplicity. And size. but yeah, I was just thinking out loud, very well could be overkill...
I've yet to try more than a few hundred words at a time in C2, although I saw a post where someone was running with a 35K word set with no worries. And I personally found C2 is near instant with partial word matches for my in game 'hints'. I have a prototype of my C2 game here.
The smaller full English dictionary (OWL) is around 220K words. A lot of word info, such as prefixes and suffixes, anagrams are separate related fields in my db, which makes it very quick. I may play around with putting this stuff in arrays in C2 and see how it runs (some of the fastest stuff I've see was done with javascript(!) ... btw, I'm by no means a pro dev, so I may not know what I'm talking about -- it's just a newly discovered hobby for now.
jobel - thanks! ... I prefer to think of it as a 'training tool' instead of cheat ...
Cheers!