py-editdist
py-editdist is a Python module to calculate the Levenshtein edit distance between two strings. It is implemented as a CPython module and is quite fast.
py-editdist is licensed under a ISC/BSD licence. See the LICENSE file for details.
The API is very simple:
import editdist # Calculate the edit distance between two strings d = editdist.distance("abc", "bcdef")
News
Mon, 28 Mar 2011: Project moved to Google Code
py-editdist has moved to Google Code. Release tarballs are available from there and should download quite a bit faster. Revision control now uses Mercurial, which supports local branches and modifications much better than CVS. Bug tracking remains at bugzilla for now, but it might too move in the future.
Moving to Google Code makes it much easier to add additional developers to the project. If you have contributed in the past to py-editdist and are interested in becoming a developer, then please let me know.
Thu, 19 Jul 2007: py-editdist-0.3 released
py-editdist-0.3 has just been released. The sole purpose of this release is to correct a bug found by Gideon Stupp that would yield incorrect return values when the edit distance between the two strings was greater than CHAR_MAX (usually 127).
Wed, 25 Apr 2007: py-editdist-0.2
I have just made a bugfix release of py-editdist to fix a bug spotted by Hamish Allan that could return incorrect results for some input cases. The release (available from the download section below) contains a tarball for Unix and binaries for Windows Python 2.4 and 2.5.
Tue, 08 Aug 2006: Windows binary available
After wrestling with Microsoft's compilers and headers, I have been able to build and release Windows binaries of the editdist module. These are available below in the download section. Some minor changes were required to the module to build it, if you are interested you can see them in CVS.
Wed, 05 Jul 2006: py-editdist-0.1 released
The first release, py-editdist-0.1.
Download
py-editdist is now hosted at Google Code. You can download tar releases or pull the current development source from there.