talkingpet

As a voice systems developers we often manually test different use cases. Testing with multiple actors could is complicated (i.e. Alice calls Bob, who add a Carol to the call)

Here is a micro-tool, based on PJSUA SIP client.

Why Talking Pet?

The name of alien from the famous computer game

Usage instructions

Build it first time:

docker-compose build --force-rm --no-cache --pull talkingpet

Configure (create an .env file from given .env-sample);
And then execute one of desired scripts:

register SIP_USERNAME -- register SIP client and let tester control the SIP client using keyboard shortcuts;
answer SIP_USERNAME -- register SIP client and automatically answer incoming calls;
play SIP_USERNAME /path/to/the/audio.wav -- register SIP client, answer incoming calls and playback given wav file;
dial SIP_USERNAME DESTINATION -- dial destination and let tester control call flow;
broadcast SIP_USERNAME DESTINATION /path/to/audio.wav -- dial destination and play audio file;

tts /path/to/the/audio Some voice announcement.

Use play and broadcast to test RTP streams and audio quality or to test call recording functionality by capturing SIP+RTP dump with sngrep utility and analysing it with Wireshark.

IMPORTANT! Audio files used by play and broadcast should be converted first to the compatible format. tts script does that or user could using ffmpeg utility (also within docker container):

ffmpeg -y -i SOURCE.wav -ar 8000 -ac 1 -ab 64K CONVERTED.wav

Examples

When inbound call will hit the conference and user on SIP/0001 answer it, announcement /recordings/line_1.wav will be played back to the calling party:

docker-compose run --rm --service-ports pjsua play 0001 /recordings/line_1.wav

Dial 00491234567890 from the trunk, and once Asterisk will answer, start playing announcement /recordings/trunk.wav:

docker-compose run --rm --service-ports pjsua broadcast trunk 00491234567890 /recordings/trunk.wav

Use both cases to test call transfers (broadcast as a trunk, receive call as an agent 0003 and transfer to the 0001).

Configuration

Configuration settings are in the .env file. IMPORTANT: except the SIP_USERNAME, it shares the same configuration options. There is two options to provide different credentials, such as a SIP password:

to copy whole folder and use different .env;

or to use -e KEY=VAL argument with docker-compose:

docker-compose run --rm --service-ports -e SIP_PASSWORD=foobar \
               pjsua play 0001 /recordings/line_1.wav

Troubleshooting

Test tool is based on Alpine linux and running in a docker, in order to troubleshoot, just enter in:

docker-compose run --rm --service-ports --entrypoint=sh pjsua

Add necessary tools such as vim or sngrep:

apk add sngrep mc vim

andrius / talkingpet