Comments on: Fun With String Searching http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/ Ye Olde Computer Science Blogge Mon, 18 Jan 2010 23:12:49 -0500 http://wordpress.org/?v=2.9.1 hourly 1 By: eran http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/comment-page-1/#comment-9905 eran Tue, 16 Jun 2009 10:57:13 +0000 http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/#comment-9905 excellent presentation ! The hardest in math is to explain things simply not to just understand them and in this you are a winner. as said before and more relevant to my case, Rabin-Karp is the best for multi-pattern searches (see wikipedia) excellent presentation !
The hardest in math is to explain things simply not to just understand them and in this you are a winner.
as said before and more relevant to my case, Rabin-Karp is the best for multi-pattern searches (see wikipedia)

]]> By: Recent Links Tagged With "rabinkarp" - JabberTags http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/comment-page-1/#comment-9852 Recent Links Tagged With "rabinkarp" - JabberTags Thu, 23 Oct 2008 10:04:35 +0000 http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/#comment-9852 [...] public links >> rabinkarp Fun With String Searching Saved by hikarisuong on Wed 22-10-2008 Rabin-Karp string search algorithm - Wikipedia, the free [...] [...] public links >> rabinkarp Fun With String Searching Saved by hikarisuong on Wed 22-10-2008 Rabin-Karp string search algorithm – Wikipedia, the free [...]

]]> By: Preslav Rachev http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/comment-page-1/#comment-9848 Preslav Rachev Sun, 19 Oct 2008 19:03:27 +0000 http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/#comment-9848 A truly great post. Thanks for the good explanation. My Data Structures professor could not explain that topic properly. Since I am mostly using C#, I tend to avoid the C/C++ way of thinking, but after I read your post, I felt really comfortable. Thanks once again A truly great post. Thanks for the good explanation. My Data Structures professor could not explain that topic properly. Since I am mostly using C#, I tend to avoid the C/C++ way of thinking, but after I read your post, I felt really comfortable.

Thanks once again

]]> By: josh http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/comment-page-1/#comment-1702 josh Thu, 13 Dec 2007 00:41:46 +0000 http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/#comment-1702 As written, Rabin-Karp is basically the linear search except you're comparing 4 characters (sort of) at a time. Without bothering to read up on it, I imagine it could win out if you're searching a fixed text for a large number of substrings. (Assuming you build an actual hash table.) IIRC both Boyer-Moore and KMP optimize the other way, for finding one string in many texts or at least with a much larger haystack than needle. That's the common case, but not the only case. As written, Rabin-Karp is basically the linear search except you’re comparing 4 characters (sort of) at a time. Without bothering to read up on it, I imagine it could win out if you’re searching a fixed text for a large number of substrings. (Assuming you build an actual hash table.)

IIRC both Boyer-Moore and KMP optimize the other way, for finding one string in many texts or at least with a much larger haystack than needle. That’s the common case, but not the only case.

]]> By: Andreas Bernauer http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/comment-page-1/#comment-1688 Andreas Bernauer Wed, 12 Dec 2007 13:51:10 +0000 http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/#comment-1688 I could improve Rabin-Karp by replacing the hash function with the simplest one: just adding all characters. Of course, this results in a lot of hash equalities where there is no match, but the extra brute_force_match calls still cost less than the overhead for the real hash value. I could improve Rabin-Karp by replacing the hash function with the simplest one: just adding all characters.

Of course, this results in a lot of hash equalities where there is no match, but the extra brute_force_match calls still cost less than the overhead for the real hash value.

]]> By: Andreas Bernauer http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/comment-page-1/#comment-1687 Andreas Bernauer Wed, 12 Dec 2007 13:02:31 +0000 http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/#comment-1687 Hi Jake, thanks for the post. I found it interesting and checked out the sources you provided to play a little bit. I found two bugs you might be interested in: - You read in the files with getline(), which discards the newline. This has two effects: (a) the reported match position is not helpful to find the string in the file. (b) Boyre-Moore isn't really "tortured", as the match is already at the beginning of the file. You can see this from your posted results: match already at position 110,154 of 11,015,556 characters. Using read() to read the file avoids both problems and also speeds up reading the large files: string result; static const int buffer_size = 1 Hi Jake, thanks for the post. I found it interesting and checked out the sources you provided to play a little bit. I found two bugs you might be interested in:

- You read in the files with getline(), which discards the newline. This has two effects: (a) the reported match position is not helpful to find the string in the file. (b) Boyre-Moore isn’t really “tortured”, as the match is already at the beginning of the file. You can see this from your posted results: match already at position 110,154 of 11,015,556 characters. Using read() to read the file avoids both problems and also speeds up reading the large files:
string result;
static const int buffer_size = 1

]]> By: Fun With String Searching : So Jake Says: http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/comment-page-1/#comment-1683 Fun With String Searching : So Jake Says: Wed, 12 Dec 2007 10:07:55 +0000 http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/#comment-1683 [...] tlipcon wrote an interesting post today onHere’s a quick excerpt [...] [...] tlipcon wrote an interesting post today onHere’s a quick excerpt [...]

]]> By: purrl.net |** URLs that purr **| http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/comment-page-1/#comment-1682 purrl.net |** URLs that purr **| Wed, 12 Dec 2007 10:06:33 +0000 http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/#comment-1682 <strong>The web's most interesting stories on Wed 12th Dec 2007...</strong> These are the web's most talked about URLs on Wed 12th Dec 2007. The current winner is... The web’s most interesting stories on Wed 12th Dec 2007…

These are the web’s most talked about URLs on Wed 12th Dec 2007. The current winner is…

]]> By: Chief http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/comment-page-1/#comment-1681 Chief Wed, 12 Dec 2007 07:58:53 +0000 http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/#comment-1681 I'd be interested to know how your algorithms test if all your multiply operations are replaced with bitshifts in Rabin-Karp. So, ALPHABET_SIZE = 256; x = ... y I’d be interested to know how your algorithms test if all your multiply operations are replaced with bitshifts in Rabin-Karp.

So,
ALPHABET_SIZE = 256;
x = … y

]]> By: Chii http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/comment-page-1/#comment-1679 Chii Wed, 12 Dec 2007 07:29:27 +0000 http://www.jakevoytko.com/blog/2007/12/11/fun-with-string-searching/#comment-1679 the third algorithm is very interesting! i wonder why it works... the third algorithm is very interesting! i wonder why it works…

]]>