UniMRCP Lex plugin parameter question

47 views
Skip to first unread message

Daniel Ng

unread,
May 14, 2021, 1:11:32 PM5/14/21
to UniMRCP
Hi all,
Sorry if this has already been covered but is there an alternative to the parameter separator?

To clarify, I'm specifying for the Voice Grammar URL:
builtin:speech/transcribe?bot-name=some_bot;alias=some_alias

The problem is that if I used the Cisco CVP Studio environment to do this, the ";" causes the CVP platform to fail a validation check (their "element" to VXML generation process looks for ";" and if it is present, checks that there are specifically 4 of them as they use the same character as separators to allow users to specify weight, language-type, etc).

Is there a setting in UniMRCP server or the Lex plugin that allows us to specify a different separator character for that URL parsing of the AWS and Lex details?  I've been through the configuration files and manuals but didn't see any, so am reaching out in case I missed something?

Kind regards,
Daniel

Vahagn Kocharyan

unread,
May 22, 2021, 12:42:14 PM5/22/21
to UniMRCP
HI Daniel
Are you tried this way?


<var name="botName" expr="'uhuhuu'"/> 
<var name="botAlias" expr="' uhuhuu  '"/> 

<grammar  srcexpr=" 'builtin:speech/transcribe?bot-name=' + botName + ';alias=' + botAlias + ';'"  />

Daniel Ng

unread,
May 24, 2021, 3:21:31 PM5/24/21
to uni...@googlegroups.com
Hi Vahagn,

Thank you for your suggestion - yes, I did try that via some static VXML but the Cisco Voice Browser didn't seem to like srcexpr (at least based on a quick test - I've still got to go hunt for the error details).  I'll have a proper look again as srcexpr is listed within the Cisco Voice Browser VXML DTD - probably user error the first time.

Either way, given the options available within CVP Studio, I suspect that I'll have to try to generate a custom VFC interface to make this user/operator friendly from a CVP Studio perspective as well as to maximise the platform's logging and reporting framework..

Kind regards,
Daniel

--
You received this message because you are subscribed to a topic in the Google Groups "UniMRCP" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/unimrcp/k3lJYoY1I8w/unsubscribe.
To unsubscribe from this group and all its topics, send an email to unimrcp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/unimrcp/c5046f43-ce37-4686-8524-d35e6ec6aab4n%40googlegroups.com.

Vahagn Kocharyan

unread,
May 24, 2021, 5:07:52 PM5/24/21
to UniMRCP
Also you can try this 

<grammar mode="voice"
       root="transcribe"
       version="1.0"
       xml:lang="en-AU"
       Fxmlns="http://www.w3.org/2001/06/grammar">
      <meta name="scope" content="builtin"/>
      <meta name="bot-name" content="asdasd"/>
      <meta name="alias" content=" asdasd  "/>
      <rule id="transcribe">
        <one-of/>
      </rule>
</grammar>
Thanks

Arsen Chaloyan

unread,
May 24, 2021, 10:04:08 PM5/24/21
to UniMRCP
Hi Daniel,

The separator can be made configurable in future versions, but you should be able to resolve the problem by using the SRGS XML grammar instead, as suggested by Vahagn. Note: a blank item might need to be specified to conform to the SRGS grammar rules.

<grammar mode="voice" root="transcribe" version="1.0" xml:lang="en-US" xmlns="http://www.w3.org/2001/06/grammar">

    <meta name="scope" content="builtin"/>
    <meta name="bot-name" content="some_bot"/>
    <meta name="alias" content="some_alias"/>
    <rule id="transcribe">
        <one-of ><item  weight='1.0' repeat='1'>blank</item></one-of>
    </rule>
</grammar>

You received this message because you are subscribed to the Google Groups "UniMRCP" group.
To unsubscribe from this group and stop receiving emails from it, send an email to unimrcp+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/unimrcp/45f5891d-8b46-4cd6-952e-328bb821dc81n%40googlegroups.com.


--
Arsen Chaloyan
Author of UniMRCP
http://www.unimrcp.org

Daniel Ng

unread,
May 25, 2021, 4:58:14 AM5/25/21
to uni...@googlegroups.com
Hi Arsen, Vahagn,

Thank you both for this - I agree that both of those mechanism should/will work at the VXML and Cisco VVB level and hence this is not a blocker for SME type custom deployments.

The organisation/company I work for tends to do hosted and custom IVR solutions and tends to prefer solutions which they can crank and churn out with the use of "standard" vendor provided software and templates where possible - hence I was looking at minimal/simplest solutions/architectures within the Cisco environment which meant the use of the Cisco CVP VXML framework (this is the Cisco's VXMLServer that generates during run-time Cisco VVB compliant VXML scripts for the Cisco VVB) where possible (and it is this Cisco VXML application server is where the Cisco platform has the issue with the ";" as the separator as their interface syntax uses that internally as their separator, causing run-time error at the VXML application server level rather than at the Voice Browser layer).

If the separator can be configurable on the Unimrcp side in future versions, this is great to know - as it means that the Cisco CVP infrastructure can then out-of-the-box, fully utilise the features offered by UniMRCP without custom development work and would be transparent to those Cisco CVP "IVR developers".  In the meantime, we will run with the suggestions provided.

thank you again.
Kind regards,
Daniel



Arsen Chaloyan

unread,
Jun 8, 2021, 9:35:30 PM6/8/21
to UniMRCP
Hi Daniel,

Thank you for your feedback. I see your point and have logged an internal ticket to possibly make the separator configurable in future versions.


Reply all
Reply to author
Forward
0 new messages