VERSION.README
1.7 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
-----------
VERSION 1.2
-----------
- VERSION 1.2 - date: April 2003
* bug fixed (April 2004):
- the phonetization of a single '0' was doubled (zero zero). This is now fixed.
* new function
- script/lia_lex2phon_variante: phonetization of a lexicon according to all the
Part-Of-Speech found in the lexicon $LIA_PHON_LEX or lex10k if not
defined AND generation of pronunciation variant according to a set of rules
contained in the file: data/rule_variante_lex.pro
- VERSION 1.1 - date: January 2003
* bug fixed:
- all 'main' function end with an 'exit(0)'
- the script 'lia_check' is done on the 'pho' format and
not the 'ola' one
* modif:
- now, the lexicon used can be specified in a envt variable:
LIA_PHON_LEX
for example: set LIA_PHON_LEX = lex10k
If this variable is not defined, the default lexicon, lex10k,
is used
- all the scripts starts now with "#!/bin/csh"
* new function
- script/lia_lex2phon: phonetization of a lexicon according to all the
Part-Of-Speech found in the lexicon $LIA_PHON_LEX or lex10k if not
defined
* new data
- a lexicon with more than 80k words is part of the distribution.
This lexicon is useful for the POS tagging programs and for the
phonetization of a lexicon. But just for text-to-speech purpose,
it's better to use only the default lex10k lexicon.
To use this lexicon, compile it first with the command:
make lex80k
Then, modify the envt variable 'LIA_PHON_LEX' as:
setenv LIA_PHON_LEX lex80k