Dummy speechd-el output driver possibly for running speechd-el inside emacs on android

[Peter Vágner]

Hello,

ESpeak does not have message priorities either, however there is a way to
interupt its speech.
Does writing speechd-el output driver require dealing with ssip?
I assumed it would be enough to wrap a few functions of the target TTS
system e.g. eSpeak implementing some functions sfeechd-el is looking for.
Since I don't know elisp at all I asked for a simplest possible output
driver for speechd-el so I can try learning more from there.
Is your direct festival support available to public so at least I might be
able to look at it what you have done at speechd-el side of things?

Thanks and greetings

Peter

D?a 4. 11. 2017 12:56 pou??vate? "Pierre Lorenzon" <
devel at pollock-nageoire.net> nap?sal:

>
> Hi Peter,
>
> From: Peter V?gner <pvdeejay at gmail.com>
> Subject: Dummy speechd-el output driver possibly for running speechd-el
> inside emacs on android
> Date: Wed, 1 Nov 2017 21:15:57 +0100
>
> > Hello,
> >
> > I know speechd-el is not in rapid development for a few years now however
> > the more I use it the more I like what I can do with emacs.
> > For a few months I'm playing with termux on android. It's possible to
> run a
> > lot of linux apps on termux for android including emacs. It's even
> > straightforward to install it without rooting or other tricks.
> > There is also a way to make linux apps to play sound through libpulse
> > running on termux and another app directly on the device.
> > Given all this I would like to try getting speechd-el working with such a
> > setup.
> > So I do have emacs, I do have speechd-el, I do have espeak built against
> > libpulse, however the glue between speechd-el and espeak is missing. I
> > can't build speech-dispatcher on termux because android pthreads support
> is
> > too limited for speech-dispatcher to build there.
>
>   In fact it would probably be good to develop a
>   speech-dispatcher version that can be compiled and run on
>   android.
>
>   Anyway I am certainly not in this mind for at least two
>   reasons : First I do not have any system with android
>   installed so no opportunity to make tests. Secondly I no
>   longer use speech-dispatcher.
>
>   In fact I developped a direct client from speechd-el to
>   festival. Since festival do not have the scheduling features
>   implemented by speech-dispatcher I had to implement a
>   scheduler for festival what I did.
>
>
> > Since speechd-el has an API that more output drivers can be running at
> the
> > same time or even independently of each other I think it might be more
> > viable than fighting pthreads on android  trying to create a kind of very
> > simple output driver for speechd-el.
> > I wish it to either interface directly with espeak running on termux or
>
>   I is certainly possible but I never use espeak so do not
>   really know what are its features. In particular is it easy
>   to interrupt a speaking message with espeak, is it possible
>   to queue the messages with given priorities. All such
>   features are needed to implement the ssip protocol. That
>   precisely the part that was badly supported by festival.
>
>   Regards
>
>   Pierre
>
>
> > there is even termux specific API that exposes native android TTS to apps
> > running on termux.
> >
> > However I don't know elisp besides some little bits I have understood
> when
> > editing emacs configuration and I don't think I might be able to move
> > forward withis idea anytime soon.
> > So my question is: might anyone else be interested in trying to make this
> > working?
> > Is there somewhere such a minimalist speechd-el output driver perhaps
> > during the development you were using it for testing, for logging or
> > similar we can get more inspiration from?
> >
> > What do you guys users as well as developers think about this?
> >
> > Sorry for taking your time and thanks for your hints.
> >
> > Greetings
> >
> > Peter
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: 
<http://lists.freebsoft.org/pipermail/speechd/attachments/20171104/d80ae82c/attachment.html>