ConversationTurn

One interaction between a human and virtual agent. The human provides some input and the virtual agent provides a response.

JSON representation 
{
  "userInput": {
    object (UserInput)
  },
  "virtualAgentOutput": {
    object (VirtualAgentOutput)
  }
}
Fields
userInput

object (UserInput)

The user input.
virtualAgentOutput

object (VirtualAgentOutput)

The virtual agent output.

UserInput

The input from the human user.

JSON representation 
{
  "input": {
    object (QueryInput)
  },
  "injectedParameters": {
    object
  },
  "isWebhookEnabled": boolean,
  "enableSentimentAnalysis": boolean
}
Fields
input

object (QueryInput)

Supports text input, event input, dtmf input in the test case.
injectedParameters

object (Struct format)

Parameters that need to be injected into the conversation during intent detection.
isWebhookEnabled

boolean

If webhooks should be allowed to trigger in response to the user utterance. Often if parameters are injected, webhooks should not be enabled.
enableSentimentAnalysis

boolean

Whether sentiment analysis is enabled.

VirtualAgentOutput

The output from the virtual agent.

JSON representation 
{
  "sessionParameters": {
    object
  },
  "differences": [
    {
      object (TestRunDifference)
    }
  ],
  "diagnosticInfo": {
    object
  },
  "triggeredIntent": {
    object (Intent)
  },
  "currentPage": {
    object (Page)
  },
  "textResponses": [
    {
      object (Text)
    }
  ],
  "status": {
    object (Status)
  }
}
Fields
sessionParameters

object (Struct format)

The session parameters available to the bot at this point.
differences[]

object (TestRunDifference)

Output only. If this is part of a [result conversation turn][TestCaseResult.conversation_turns], the list of differences between the original run and the replay for this output, if any.
diagnosticInfo

object (Struct format)

Required. Input only. The diagnostic [info][Session.DetectIntentResponse.QueryResult.diagnostic_info] output for the turn. Required to calculate the testing coverage.
triggeredIntent

object (Intent)

The Intent that triggered the response. Only name and displayName will be set.
currentPage

object (Page)

The Page on which the utterance was spoken. Only name and displayName will be set.
textResponses[]

object (Text)

The text responses from the agent for the turn.
status

object (Status)

Response error from the agent in the test result. If set, other output is empty.

TestRunDifference

The description of differences between original and replayed agent output.

JSON representation 
{
  "type": enum (DiffType),
  "description": string
}
Fields
type

enum (DiffType)

The type of diff.
description

string

A human readable description of the diff, showing the actual output vs expected output.

DiffType

What part of the message replay differs from the test case.

Enums
DIFF_TYPE_UNSPECIFIED Should never be used.
INTENT The intent.
PAGE The page.
PARAMETERS The parameters.
UTTERANCE The message utterance.
FLOW The flow.