[asterisk-users] Synchronous dialplan execution for feedback while processing speech recognition and voice synth, for example.

[Jonathan H]

I've got an agi that recognises speech (via Google) and another that turns
text into speech (tts) (via Microsoft Translate).

Both are web APIs, both called via seperate python AGIs.

I've googled and I'm probably missing something pretty newbie 101 here, but
is there any way, or fiddle, that I can play some audio to let the caller
know that their weather forecast is being fetched, which the two agis are
first recognising and then synthesizing the speech? I thought "background"
might do it, but it doesn't seem to work like that.

Both agis take about 2.5 seconds, so that's 5 seconds of silence.

Yes, I'm aware of software applications that will quickly recognise and
synthesize speech on the local machine, but they're not suitable for the
application.

I realise there's probably a pythonic way of forking processes, but the way
it is now, both agis are nice seperate single purpose functions and it all
just looks quite neat as it is.

Many thanks.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.digium.com/pipermail/asterisk-users/attachments/20161003/f8fdf48a/attachment.html>