Embeddable Predictive Text Library: lib378 (C) & lib378js (Javascript)

Overview

This library provides a simple C API for partial and full string matching, it can provide predictive text functionality like T9(C) and similar.
Dictionary is stored in a highly optimized compressed file format (bzip2-like performance algorithm) which can be optionally further bzip2 compressed to achieve really minimum space requirements. There's also an experimental but quite usable JavaScript implementation of the algorithm that can be easily integrated into browsers/web pages.

See this 245239 words comparison:

• Plain ASCII (sorted) download [2.6M]
• Custom file format download [744K] (~3bytes/word, close to bzip2 performance!)
• Bzipped2 custom file format download [200K] (~1byte/word with optional bzip2 over previous one!)

Screenshots

A rather unexciting screenshot of the example.

Download

Snapshots/Releases

Grab latest snapshot or release from project's main page http://sourceforge.net/projects/lib378/

Anonymous CVS:

This is the preferred way to test all the latest features and fixes.

	cvs -d:pserver:anonymous@lib378.cvs.sourceforge.net:/cvsroot/lib378 login
 
	cvs -z3 -d:pserver:anonymous@lib378.cvs.sourceforge.net:/cvsroot/lib378 co -P lib378

Example

For a sample 9-keys implementation see the example.

License

GNU Library or Lesser General Public License (LGPL)

Copyright/Author

(C) 2007-2010 Massmiliano Galanti, <massiblue@libero.it>

Links

• Sourceforge project page: http://sourceforge.net/projects/lib378/
• Freshmeat project page: http://freshmeat.net/projects/lib378/
• lib378.js JavaScript implementation: demo

---