Hello @yleny, its a bit confusing at the beginning. I guess youâre confortable with the linux command line, so first you will need to find the proper mycroft.conf. I did found the correct one was the placed on /etc/mycroft/mycroft.conf
(that will persists the core upgrades as fas as I can see, but that file is almost empty, so you should take a look into /usr/local/lib/python2.7/site-packages/mycroft_core-0.9.8-py2.7.egg/mycroft/configuration/mycroft.conf
(perhaps your mycroft core version may be different, it seems Iâm on the latest version, I donât know why.
On that last mycroft.conf you will find a complete and more comprehensive configuration that can be copy-pasted to /etc/mycroft/mycroft.conf
and modified at will. I will post mine here (I have mycroft configured to âhearâ in spanish)
{
"enclosure": {
"platform": "mycroft_mark_1",
"platform_build": 10,
"port": "/dev/ttyAMA0",
"rate": 9600,
"timeout": 5.0,
"update": true,
"test": false
},
"VolumeSkill": {
"default_level": 6,
"min_volume": 0,
"max_volume": 83
},
"ipc_path": "/ramdisk/mycroft/ipc/",
"lang": "es-es",
"tts": {
// Engine. Options: "mimic", "google", "marytts", "fatts", "espeak", "spdsay"
"module": "espeak",
"mimic": {
"voice": "ap"
},
"espeak": {
"lang": "es-es",
"voice": "m1"
}
}
}
After rebooting the Mark1 (or restarting the daemons) Mycroft understands spanish. Now you have to translate the skills into your language. The way Iâm doing it is as explained in this thread, go where skills are placed and duplicate the directory en-us into your language code (mine is es-es) for example:
dialog/en-us/speak.dialog
vocab/en-us/sentence.intent
vocab/en-us/keyword.voc
regex/en-us/regex.rx
Then open an editor and translate all the sentences, verbs and everything you will find into you language.
After rebooting (or restarting the daemons) mycroft should understand to you and answer in your own language.
Last step is to translate the default phrase Mycroft says when it doesnât understands you, but itâs not yet ready. Itâs hardcoded in the core, on /usr/local/lib/python2.7/site-packages/mycroft_core-0.9.8-py2.7.egg/mycroft/client/speech/main.py
(your core version may differ) and search for the sencence âSorry, I didnât catch that. Please rephrase your request.â
def handle_complete_intent_failure(event):
LOG.info("Failed to find intent.")
# TODO: Localize
data = {'utterance':
"Sorry, I didn't catch that. Please rephrase your request."}
ws.emit(Message('speak', data))
As you can see, there is a TODO:Localize, so letâs hope Mycroft devs shows some love to non-english users and let us translate this and others into our languages
If you translate this default phrase, you will lost it when will be an update of the core.
I do not have STT configured on my modified configuration, but taking a look to mycroft.conf at core directory, itâs already configured for mycroft
"stt": {
// Engine. Options: "mycroft", "google", "wit", "ibm", "kaldi"
"module": "mycroft"
// "kaldi": {
// "uri": "http://localhost:8080/client/dynamic/recognize"
// }
}
Hope it helps.