This is also a rather big contribution. Most of it comes from my old (2010) contributions to the festival Debian package.
On the language selection perspective, instead of a hardcoded list of options in language.list, we use a system similar to the one used in voices (proclaim_language) which allows third party language modules to define themselves by placing a file in the libdir/languages/ directory.
On the voice selection perspective, a voice.find function is provided that enables us to find installed voices that fulfill a criteria (such as the gender, language or engine that they use if they specify it).
The default voice list is extended to include on the lowest priority any installed voices. The aim is that even if the user did not install any of the hardcoded default voice, a voice is always selected.
Last but not least, a few global synthesis options were missing in the voice_reset function. I added them.
If you think this is too much, feel free to tell me. In Debian it has been useful...
This is also a rather big contribution. Most of it comes from my old (2010) contributions to the festival Debian package.
On the language selection perspective, instead of a hardcoded list of options in
language.list
, we use a system similar to the one used in voices (proclaim_language) which allows third party language modules to define themselves by placing a file in the libdir/languages/ directory.On the voice selection perspective, a
voice.find
function is provided that enables us to find installed voices that fulfill a criteria (such as the gender, language or engine that they use if they specify it).The default voice list is extended to include on the lowest priority any installed voices. The aim is that even if the user did not install any of the hardcoded default voice, a voice is always selected.
Last but not least, a few global synthesis options were missing in the
voice_reset
function. I added them.If you think this is too much, feel free to tell me. In Debian it has been useful...