U++ forum: Welcome to the forum

Re: Substring search algorithm [message #35535 is a reply to message #35532]

Mon, 27 February 2012 10:55

mr_ped
Messages: 826
Registered: November 2005
Location: Czech Republic - Praha

Experienced Contributor

"Boyer-Moore-Horspool" and family of those are suitable for U++ implementation (I did it years ago in Pascal at uni), but I wonder what's the real speed up benefit in real world app, because nowadays the CPU + L1 cache operates light years faster than L2+ cache/RAM, so while BMH will save you some compares, it has to fetch the full text anyway, and I would expect a properly implemented naive byte compare will easily do it's work inside the "time window" of memory fetch of next data.

But I may try to write BMH for U++ and toy around with that for a while to see some real numbers.

[Updated on: Mon, 27 February 2012 10:55]

Report message to a moderator