load multiple grammar

12 Jan 2017

      Hi,

I need help regarding loading and activating multiple grammars.

Per FS-2906<https://freeswitch.org/jira/browse/FS-2906>, it is possible to load multiple grammars from FreeSwitch.

I have tried the same using following commands from my C# code. The FS version I have is 1.3.17.

InboundSession.Execute("detect_speech", "unimrcp {start-recognize=false}yesnod yesnod");
InboundSession.Execute("detect_speech", "unimrcp {start-recognize=true}yesno yesno");

I can see that there are two DEFINE-GRAMMAR requests sent to the LumenVox Speech Server. One with yesnod.gram and another with yesno.gram
But there is only one session mentioned in the RECOGNIZE request where as I am expecting it to be having both the sessions described in the DEFINE-GRAMMAR request.
Any idea?

                MRCP/2.0 226 RECOGNIZE 3
                Channel-Identifier: 115483E487508175CFDF@speechrecog
                Content-Type: text/uri-list
                Cancel-If-Queue: false
                Start-Input-Timers: false
                Speech-Complete-Timeout: 3000
                Content-Length: 13

                session:yesno
...
...
...
...
There should be session:yesnod also.
I do not see a way to do this in detect_speech.

Any help would be great.

Here is the define grammar message too...

                MRCP/2.0 318 DEFINE-GRAMMAR 1
                Channel-Identifier: 115483E487508175CFDF@speechrecog
                Content-Type: application/srgs
                Content-Id: yesnod
                Content-Length: 158

                #ABNF 1.0 UTF-8;
language en-US;
mode dtmf;
tag-format <semantics/1.0>;

root $yesorno;

$yes = 1 {out = "yes"};
$no = 2 {out = "no"};
$yesorno = $yes | $no;

01/12/2017 07:16:09.790,RECV,OnLoadGrammarRe,Received cseq grammar load (1) reply for CallIndex 51
01/12/2017 07:16:09.790,SEND,SendPacket     ,

                MRCP/2.0 116 1 200 COMPLETE
                Channel-Identifier: 115483E487508175CFDF@speechrecog
                Completion-Cause: 000 success

01/12/2017 07:16:10.123,RECV,ProcessRcvdPkt ,Got MRCP Message -

                MRCP/2.0 408 DEFINE-GRAMMAR 2
                Channel-Identifier: 115483E487508175CFDF@speechrecog
                Content-Type: application/srgs
                Content-Id: yesno
                Content-Length: 249

                #ABNF 1.0 UTF-8;
                language en-US; //use the American English pronunciation dictionary.
                mode voice;  //the input for this grammar will be spoken words.
                tag-format <semantics/1.0>;

                root $yesorno;

                $yes = yes;
                $no = no;
                $yesorno = $yes | $no;

01/12/2017 07:16:10.128,RECV,OnLoadGrammarRe,Received cseq grammar load (2) reply for CallIndex 51
01/12/2017 07:16:10.129,SEND,SendPacket     ,

                MRCP/2.0 116 2 200 COMPLETE
                Channel-Identifier: 115483E487508175CFDF@speechrecog
                Completion-Cause: 000 success

01/12/2017 07:16:10.459,RECV,ProcessRcvdPkt ,Got MRCP Message -

                MRCP/2.0 226 RECOGNIZE 3
                Channel-Identifier: 115483E487508175CFDF@speechrecog
                Content-Type: text/uri-list
                Cancel-If-Queue: false
                Start-Input-Timers: false
                Speech-Complete-Timeout: 3000
                Content-Length: 13

                session:yesno

01/12/2017 07:16:10.460,DEBG,SetRecPrmFrmHdr,Setting parameter 'Speech-Complete-Timeout' with value '3000'
01/12/2017 07:16:10.460,DEBG,SetRecPrmFrmHdr,Setting parameter 'Start-Input-Timers' with value 'false'
01/12/2017 07:16:10.460,DEBG,ConfPortForReco,Setting speech_complete_timeout to 3000 ms
01/12/2017 07:16:10.460,DEBG,ConfPortForReco,Setting sensitivity level to 50
01/12/2017 07:16:10.460,DEBG,ConfPortForReco,Setting Speed vs Accuracy to 0.5
01/12/2017 07:16:10.461,SEND,SendPacket     ,

                MRCP/2.0 87 3 200 IN-PROGRESS

Thanks and regards,
Nagendra

Nagendra Babu

Nagendra Babu

Michael Jerris

tags

participants (2)