At present MaryTTS nodes can be instructed to continue the state machine immediately while speech is still ongoing. Unfortunately, there's no way to stop speaking from the state machine.
I wonder, @alexanderkoller , @akoehn , if you have opinions on how this should be implemented. As a node, as a script action (given that this will most likely be performed in relation to some other script stuff)? Ideas?
solution: I've decided to scrap the "feature" that two syntheses can be ongoing at the same time. That way, a second empty synthesis (e.g. of " ") will abort any synthesis that is still ongoing. See 6fc6248
At present MaryTTS nodes can be instructed to continue the state machine immediately while speech is still ongoing. Unfortunately, there's no way to stop speaking from the state machine.
I wonder, @alexanderkoller , @akoehn , if you have opinions on how this should be implemented. As a node, as a script action (given that this will most likely be performed in relation to some other script stuff)? Ideas?