This code is no longer maintained by me. Check out
http://www.darrensessions.com/docs/app_swift/
or
http://sourceforge.net/projects/appswift/
for newer versions!
Cepstral Swift is a low-cost text-to-speech engine. You need to have this installed and working before trying to integrate it into Asterisk.
Several people have written applications to glue Swift into Asterisk:
These all can be made to work, but they all bugged me in one way or another. I wanted code that:
- Doesn't keep the caller waiting in silence while the app generates the entire TTS output to a temp file
- Doesn't unceremoniously kill off the swift engine when done, upsetting the Cepstral licesnse server and eating a concurrency license
- Has configurable in-memory buffering of the swift output to balance memeory usage vs Swift process concurrency
- Responds to user DTMF during the speech by setting a channel variable and optionally doing a goto of the extension entered
So I wrote a version of app_swift addresses all that. It can be downloaded here:
app_swift-0.9.tar.gz for Asterisk 1.2 (no longer maintained)
app_swift-0.9.1.tar.gz for Asterisk 1.4
Unpack, make, make install. It works for me on FC4 Linux/Asterisk 1.2.10 and Asterisk 1.4B3. Feel free to email me comments, updates, fixes, or problems. Thanks!
I release the above code into the public domain. I do not have time to update for new versions of
asterisk or to provide new features (as of Apr 2008). Feel free to redistribute with your own changes
or I can post patches/new revs here if anyone wants to...
will@loopfree.net