Platform Release Version 7.1.3

[CP-5631] - Fixed: Crawler - Test Case generation with reserved characters

Attempting to generate Test Cases from a Model, with reserved characters in the folder name, would cause an "Internal Error" popup to be displayed, and Test Case generation would fail. Additional validation has been added to handle this scenario.

[CP-5819] - Fixed: Crawler - Test Case Generation with duplicate Inputs

If a Node in a Crawler Model contained two or more child Nodes with the same Input, the additional Nodes were considered duplicates and silently not included when generating Test Cases. Additionally, the validation failure message "There is nothing to export" was displayed when attempting to generate Test Cases from a duplicate Node directly. Duplicate Nodes are now included when generating Test Cases.

[CP-5938] - Fixed: Crawler - Out of Memory Exception

When Crawling very deep IVRs, it was possible for Crawls to fail due to an Out of Memory exception.

[CP-6025] - Fixed: In-progress Campaign Run Summary Report

Filtering the Campaign Summary Report of an in-progress Campaign Run, e.g. selecting only Satisfactory Results, could cause an "Internal Server Error" popup to appear.

[CP-6053] - Fixed: Crawler - Node update after toggling Allow Crawl

When toggling the "Allow Crawl" setting for a Node, the Model view was incorrectly not displaying the change unless the page was refreshed.

[CP-6054] - Fixed: Crawler - Menu Name Change

The Menu Name for a Node was incorrectly being changed to "Not Crawled" if the "Allow Crawl" setting for that Node was set to one of the "No" options.

[CP-6062] - Fixed: Call Engine Stack Trace in Distributed ASR mode

If the CE was configured to use Distributed ASR, when a Call was aborted before all recognition requests had been fulfilled, an error and stacktrace were logged for each recognition not completed.

[CP-6077] - Fixed: Crawler - Node modified manually and by Crawler

If the following conditions were met:
1) A Crawl was running for a Model
2) A Node was manually modified for the same Model, while the Crawl was still running
3) The Node was attempted to be saved (either automatically or by clicking away from the form) Then a modification validation message could be displayed to the User, requiring the Node to be reloaded before any changes could be made, which would subsequently result in the unsaved changes being discarded. This message will now only be displayed if the same Node is both Crawled and edited manually by the User, the Node has not been marked as User Verified, and the Prompt is mismatched.

[CP-6083] - Fixed: Prompt Analyzer Distributed ASR

If the Portal was configured to use Distributed ASR, and a Prompt Analyzer speech recognition request timed out, an error and stacktrace were logged in the Portal's log file.

[CP-6104] - Fixed: Crawler - Modifying generated Blocks

Blocks generated from a Crawler Model were unable to be modified in their original location, as the Web Portal was failing to load the Block's folder path.

[CP-6112] - Fixed: CVA - Error using reason codes for Cisco Finesse

After downloading the list of Reason Codes for a given user from Cisco Finesse, CVA could incorrectly handle the response if it contained more than one Reason Code. This would result in an Internal Error result for a Campaign Run if Agents were configured to use Reason Codes.

[CP-6134] - Fixed: Discrepancy in Test Step and Prompt Analyzer confidence scores

When using Distributed ASR, a discrepancy in the confidence scores for a Test Case Step, between running the Step via Prompt Analyzer versus running the Step as part of a Test Case (in a Campaign Run or Test Case Validation), has been fixed.

[CP-6141] - Fixed: Call Engine Cruncher Lite Mode

If the Call Engine was configured in Cruncher Lite mode, and was also configured to use Distributed ASR, Test Results would incorrectly be marked as failed with "Expected n steps but got 1", instead of all Test Step Results requiring recognition being marked as 100% success.

[CP-6163] - Fixed: Voice Scheduler Database Deadlock

In some scenarios, the Voice Scheduler could encounter a deadlock situation while simultaneously concluding a Campaign Run and inserting a new Voice Test Result into the database.

Was this article helpful?

0 out of 0 found this helpful