Boost logo

Boost :

Subject: [boost] [algorithm] Review Request: edit_distance
From: Erik Erlandson (eje_at_[hidden])
Date: 2014-01-27 12:26:13


I would like to request a formal review for inclusion of the edit_distance() function.

Description:
The edit distance is the length of the shortest (or least-cost) edit script from one sequence to another, where an edit script is defined as a sequence of insertion, deletion and (optionally) substitution operations. The function implementing the edit distance is named edit_distance. This function will return the edit distance between two sequences, where sequences may be any valid range object supporting forward iteration. The edit_distance function will also, if requested, return the edit script.
http://en.wikipedia.org/wiki/Edit_distance

Documentation:
http://erikerlandson.github.io/algorithm/libs/algorithm/doc/html/algorithm/Sequence.html

Code resides on a branch of my fork of Boost.Algorithm:
https://github.com/erikerlandson/algorithm/tree/edit_distance

The feature branch above includes the doc, unit tests and examples that are integrated via Jamfiles. The new dev resides specifically in this subdir:
https://github.com/erikerlandson/algorithm/tree/edit_distance/sequence

I have built the tests and examples on gcc and clang, with and without c++11 mode. I don't have access to a windows/MSVC build env, so I am definitely curious if it compiles properly in that env.

My concept is that boost::algorithm::sequence can eventually serve as a home to additional related algorithms, for example other sequence comparison algorithms or perhaps other categories of sequence-related algorithms.


Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk