Hey there,
When running a test case where the bot has two responses (one specific to the utterance, the other a generic “for more information” response), the test case will only pass if the first response is in the #bot.
for example:
#me
a question
#bot
an answer
a generic answer
The bot i’m connecting to is a lex bot. it seems that when two responses are separated out on the UI, it’s not simply a matter of just having the two responses together in the .convo text file.
Has anyone every come across this before? i’ve pasted a screenshot of the bot’s UI to show what I mean by two separate responses
Apologies if this post isn’t clear
Can be solved by using two #bot sections instead of one
#me
a question
#bot
an answer
#bot
a generic answer
Hey Florian,
Thank you for replying. I’ve tried this approach but I’m afraid the console claims that the bot did not respond within 10 seconds. I’m very new to Botium, perhaps this is an issue with Lex rather than botium? I’ve included a screenshot of my test case (with an attempt to add a pause to). I’m not sure about confidentiality so I just went on the safe side and blocked out the client’s information.
No worries, if you’re unable to answer this query. I will find some sort of workaround if needs be.
I must say, Botium is such a fabulous tool.
When Lex is returning multiple result bubbles then Botium should reflect this.
Can you pls run it with verbose output (–verbose if using botium-cli, export DEBUG=botium* for botium-bindings) and post the result here - and of course yes, you should mask confidential information.
