NAME

Text::Levenshtein::Damerau::XS - XS Damerau Levenshtein edit distance.

SYNOPSIS

        use Text::Levenshtein::Damerau::XS qw/xs_edistance/;

        print xs_edistance('Neil','Niel');
        # prints 1

DESCRIPTION

Returns the true Damerau Levenshtein edit distance of strings with adjacent transpositions. XS implementation (requires a C compiler). Works correctly with utf8.

        use Text::Levenshtein::Damerau::XS qw/xs_edistance/;
        use utf8;

        xs_edistance('ⓕⓞⓤⓡ','ⓕⓤⓞⓡ'), 
        # prints 1

Speed improvements over Text::Levenshtein::Damerau::PP:

        # Text::Levenshtein::Damerau::PP::pp_edistance("four","fuor")
        timethis 1000000: 381 wallclock secs (380.45 usr +  0.01 sys = 
                 380.46 CPU) @ 2628.40/s (n=1000000)

        # Text::Levenshtein::Damerau::XS::xs_edistance("four","fuor")
        timethis 1000000: 19 wallclock secs (19.43 usr +  0.00 sys = 
                 19.43 CPU) @ 51466.80/s (n=1000000)

METHODS

xs_edistance

Arguments: source string and target string.

  • OPTIONAL 3rd argument int (max distance; only return results with $int distance or less). 0 = unlimited. Default = 0.

Returns: int that represents the edit distance between the two argument. Stops calculations and returns -1 if max distance is set and reached.

Wrapper function to take the edit distance between a source and target string using XS algorithm implementation.

        use Text::Levenshtein::Damerau::XS qw/xs_edistance/;
        print xs_edistance('Neil','Niel');
        # prints 1

       # Max edit distance of 1
        print xs_edistance('Neil','Niely',1); # distance is 2
        # prints -1

TODO

  • Handle very large strings of text. Can be accomplished by reworking the scoring matrix or writing to disk.

  • Add from_file methods.

  • Add binary/byte string support.

SEE ALSO

BUGS

Please report bugs to:

https://rt.cpan.org/Public/Dist/Display.html?Name=Text-Levenshtein-Damerau-XS

AUTHOR

Nick Logan <ugexe@cpan.org>

LICENSE AND COPYRIGHT

This library is free software; you can redistribute it and/or modify it under the same terms as Perl itself.