Re: [boost] [gsoc 2013] Approximate string matching

Subject: Re: [boost] [gsoc 2013] Approximate string matching
From: Jan Strnad (hanny.strnad_at_[hidden])
Date: 2013-04-29 07:18:52

Dne 29.4.2013 03:07, Erik Erlandson napsal(a):
>
> I have been wondering if these distance functions should be specifically part of string algorithms, or more general algorithms.
>
> For example, edit distance is really a function on any two sequences, not just sequences of characters.   Rather like std::sort() is an algorithm that can be applied to a sequence of any kind of object.
>

Well, I depends on the distance itself. Hamming, Levenstein and their 
relatives can be used with any kind of object (as long as it implements 
proper == operator).

To compute Delta and Gamma distances I must be able to compute distance 
between two non-equal objects. This is easy for characters, but may not 
be so easy for "regular" objects.

I'm not sure about Jaro and Jaro-Winkler distances right now, but I 
believe it is the same case as Hamming distance -- generalization should 
be possible.

Hope this answers your question.

Jan Strnad