(Page 3 of 3 in this chapter)

5.2 Demonstration Programs

Program	Description
`ttstest`	NaturalText interactive function demonstration.
`intts`	Inbound call demonstration.

5.2.1 Interactive Test Program: ttstest

Tests all NaturalText functions. ttstest is used to: Verify proper installation and operation of NaturalText. Allow users to experiment with NaturalText and observe its dynamics.

ttstest [options]
where the options are: Option Description -? Help -b n AG board number <n>. Default=0. -s n:m Specifies the MVIP stream and timeslot for the first channel. Default = 0:0 -p protocol Specifies the protocol to run. Default = nocc. -f filename Name of file to speak. Program loops to speak specified file continuously until stopped by the user. -m method E-mail preprocessor method 0 = keep header 1 = strip header 2 = none (default) -v v Speaking voice. Default = 1. 1 = English male 2 = Spanish male

Option	Description
-?	Help
`-b` n	AG board number <n>. Default=0.
`-s` n`:`m	Specifies the MVIP stream and timeslot for the first channel. Default = 0:0
`-p` protocol	Specifies the protocol to run. Default = nocc.
`-f` filename	Name of file to speak. Program loops to speak specified file continuously until stopped by the user.
-m method	E-mail preprocessor method 0 = keep header 1 = strip header 2 = none (default)
-v v	Speaking voice. Default = 1. 1 = English male 2 = Spanish male

ttstest is a menu-driven interactive program. Enter one- and two-lettered mnemonics to execute NaturalText commands. Some commands prompt for additional information, such as file names or text strings.
You can enter commands while speech is in progress. For example, while NaturalText is speaking, you can run speech control functions such as modify speed (MS).
To initiate multiple instances of ttstest simultaneously, assign each instance a different stream and timeslot number by using the command line option `-s'.

To run ttstest, follow the steps listed below: 1. From the command line prompt, run ttstest (with or without options): ttstest [-b n -s n:m -p protocol -f filename] If you run ttstest with the -f option, it continuously speaks the text in the specified file until you interrupt it by entering either the SS or Q command. 2. If you run ttstest without the -f options, the program provides a list of all ttstest commands (see Notes). You cab run a typical test with the following commands : OP open context. Opens a context on the board, stream, and slot (defaults are 0,0, and 0). SP start protocol. The default protocol is nocc (no call control protocol). OT open NaturalText text-to-speech conversion service (TTS). IF preprocess text in specified file SF run text-to-speech conversion of specified file S? display process status of TTS CT close TTS service CP close context on which TTS is running Q end ttstest

When you run ttstest, the program displays the following list of commands: H help Q quit ! repeat command OP open port CP close port X show port state SP start protocol UP stop protocol AC answer call RC release call C? call status IF file preproc. IC cont preproc. OT open TTS CT close TTS SF speak file SI speak user input S? speak status SS speak stop P speak pause SR speak resume SJ speak jump MG modify gain MP modify pitch MS modify speed DL link a UED DU unlink a UED

Option	Description
`-?`	Help
`-b` n	Specifies the AG board number <n>. Default = 0.
`-s` n:m	Specifies the MVIP stream and timeslot for the first channel. Default = 0:0
`-p` protocol	Specifies the protocol to run. Default = did0.
`-l`	Shows low-level events.
`-v` voice	Speaking voice. Default = 1. 1 = English male 2 = Spanish male

ttsGetStatus, ttsJump, ttsPause, ttsResume, ttsSetGain, ttsSetPitch, ttsSetSpeed, ttsSpeak, ttsStop

This description applies to the wink-start protocol did0.
When you take the handset off-hook, the TCP acknowledges the line seizure by generating a wink. After the wink, enter three DTMF digits as the address. Depending upon the first digit entered, you should hear a reorder tone, busy tone or ringing. If the call is connected, the demonstration proceeds into the conversation state.
Once in the conversation state, you are prompted for one of the following options: Listen to e-mail Hang up the line

The following procedure assumes that you are testing on an AG-8 with a loop-start hybrid on line 0.
To run intts: 1. Start intts by entering the following at the prompt: intts -b n -s n:m -p lps0-v v Specify the AG board and MVIP stream for the handset you are using. 2. Put the telephone off-hook. You should hear a relay click. Dial a three digit telephone number. The program plays a prompt that explains the option of listening to e-mail or hanging up. 3. Select the "listen to e-mail" option. You will hear a sample e-mail spoken by NaturalText. The e-mail includes instructions for using a touch-tone phone to control the speech. Here is a list of the controls available: DTMF Digit Control 3 jump forward 6 jump back 1 speak louder 2 speak softer 4 speak faster 5 speak slower 7 increase pitch 8 decrease pitch * pause # resume 0 stop 4. When speaking is finished, you are returned to the prompt to listen or hang up. 5. Select the hang-up option. The line hangs up and the demo waits for a new inbound call.

intts demonstrates synchronous programming on a single port. Many of the demonstration subroutines are coded this way, such that a "start" command is initiated and the demonstration continuously retrieves events from the AG board until the expected event returns. This simplifies the overall program flow.
This demonstration can be used with various protocols or configurations. If no call control is used (protocol = nocc), the demonstration code does not wait for a call. Instead, it begins playing the prompt.

(Page 3 of 3 in this chapter)