README for the Italian voices of Festival
=========================================

Encoding of accented letters
----------------------------

The italian database uses the Latin1 encoding for accented characters.
If input text encoded in UTF8 fails like this:

  $ echo caffè | festival --language italian --tts
  11:49 <riccardo> LTS_Ruleset italian_downcase: no rule matches:
  11:49 <riccardo> LTS_Ruleset: # c a f f *here* \uffff \uffff #

then you need to recode the input.  Just put recode in the pipe:

  echo caffè | recode utf8..lat1 | festival --language italian --tts

You can query the wanted encoding using this LISP code:

  For the Italian female voice:
  (or (cadr (assoc 'coding (cadr (voice.description 'lp_diphone)))) 'ISO-8859-1)

  For the Italian male voice:
  (or (cadr (assoc 'coding (cadr (voice.description 'pc_diphone)))) 'ISO-8859-1)

  For whatever is the current voice:
  (or (cadr (assoc 'coding (cadr (voice.description current-voice)))) 'ISO-8859-1)