in reply to finding longest common substring
Without looking at it too closely, that looks fairly clean. It isn't terribly efficient though. It won't scale well to very large strings, especially if the common substrings end up being relatively small. In the worst case, you try (N^2+N)/2 substrings. (Try that with ['a' . 'X' x 10000, 'a' . 'O' x 10000] for instance...)
If you were doing this in C, you'd probably use Suffix Trees (offsite). Someone once asked about those at (the poorly named) suffix arrays but nothing much came of it except a suggestion by Zaxo that substr() could be used to create a data structure similar to a suffix tree and a note by BrowserUk that Zaxo's suggestions wouldn't work with versions less than 5.8.1.
Update: Looking at Dominus's site, it turns out that the discussion on Expert QOTW #14 referenced by others also mentions suffix trees. Dominus points to this explanation of them. Warning: that link is to a directory with a bunch of stuff in it. This might be a better starting point. Or, if you prefer, the PDF version. There's also a powerpoint version in there.
-sauoq "My two cents aren't worth a dime.";
|
|---|