This is really cool, but I'm curious what the performance implications are when doing fuzzy matching? I noticed that the list of types of errors did not include transposition. I wonder if the reason for this comes down to performance/technical reasons.
He talks about the time complexity ("O(M^2N), where M is the length of the regular expression and N is the length of the text"), and he talks more about it under "Predictable matching speed." It doesn't look like he talks about how the fuzziness feature affects that but I could have missed it.
Adding transposition I've thought about, and received a number of requests for, but never got around to doing it. The reason isn't performance, but doing a change like this would get quite involved with the guts of the library, and I simply don't have the time...
[+] [-] rplnt|14 years ago|reply
1. https://github.com/seatgeek/fuzzywuzzy
[+] [-] schwa|14 years ago|reply
[+] [-] Wilduck|14 years ago|reply
[+] [-] coderdude|14 years ago|reply
He talks about the time complexity ("O(M^2N), where M is the length of the regular expression and N is the length of the text"), and he talks more about it under "Predictable matching speed." It doesn't look like he talks about how the fuzziness feature affects that but I could have missed it.
[+] [-] metafunctor|14 years ago|reply