TODO

TODO

* extend main algorith to support more flexible string metaphonization,
once names mostly will be source of rules exception. This should result
into three types of functions:

  1) STRICT to phonetics rules. No exceptions beyond that known on Portuguese

  2) LESS-STRICT, where some common exceptions are included. Those mostly 
     exists to solve foreign names and strange names. Only common cases are
     supported, so many exceptions will follow one or another rule, even that
     if that means to map some unusual sound used by people to another key.

  3) STRICTLESS. Not certain if this level should be reached. However, it 
     will be less strict, possibly omitting any phonetic key that could make 
     search more difficult. Mostly, keeps attached to sound representation, 
     more than orthografy rules, which it means considering the sound use of
     the letters and words.
     
* add support to PHP
* add support to MySQL