DamerauLevenshteinDerivor

Author: Amber & Zooko Wilcox-O'Hearn

Contact: amber@cs.toronto.edu, zooko@zooko.com

Released under the GNU AFFERO GENERAL PUBLIC LICENSE, see COPYING file for details.

Functionality in Brief:

DamerauLevenshteinDerivor reads a vocabulary file that has one word per line in it, storing the words into a Judy array. The vocabulary can then be queried. In addition, derivor will return a list of all words in the vocabulary that are of Damerau-Levenshtein distance one from a given word.

The storage and calculations are done in C and available through the C executable, as well as through a Python interface.

DamerauLevenshteinDerivor requires:

A C compiler, Python, make, Judy Trees library (http://judy.sourceforge.net/)

It was tested under the following versions:

• Judy Trees 1.0.5
• Ubuntu 12.04.2 LTS
• Python 2.7.3
• GNU make 3.81

run :

$ make

run :

$ make test

run :

$ make install

This will copy the derivor executable into $prefix/bin and will execute python setup.py install, which will result in the derivor Python module being copied into the appropriate place in your Python installation.

>>> from DamerauLevenshteinDerivor import cderivor >>> d = cderivor.Derivor('my_vocabulary_file') >>> d.variations('as') ['a', 'is', 'us', 'al', 'an', 'at', 'has', 'was'] >>> d.inV('as') True >>> d.inV('asx') False