Support For Unicode 3.0 May Cause Map Validation Errors In TIBCO Scribe® Online

Support For Unicode 3.0 May Cause Map Validation Errors In TIBCO Scribe® Online

book

Article ID: KB0078211

calendar_today

Updated On:

Products Versions
TIBCO Cloud Integration - Connect ( Scribe ) V2

Description

Issue: Prior to the Summer 2017 release, if there were letters from other languages, such as Cyrillic characters, in Entity or Field Names in the datastore or you were using Unicode characters from a version of Unicode higher than 3.0,  TIBCO Scribe® Online stripped out anything that was not a-z, A-Z, 0-9, and _. 


After the Summer 2017 release, Unicode up to version 3 is supported for Entity and Field Names.  Due to the original method used by TIBCO Scribe® Online to strip out unsupported characters, resetting metadata, and editing/saving a Map after the Summer 2017 release of TIBCO Scribe® Online may cause Map validation errors. (42288)


NOTE: Connection names, user names, and passwords still do not support anything except a-z, A-Z, 0-9, and _ .


If a TIBCO Scribe® Online Organization has entities or field names that have been scrubbed by the stricter method, limiting everything to a-z, A-Z, 0-9, and _, the following rules apply as long as metadata is not reset:
 
  • Agents can upgrade
  • Maps continue to run after the changes
  • Maps can be edited


Resetting Metadata

If you reset metadata the following rules apply:
  • Maps continue to run.
  • Editing Maps triggers validation errors.
  • Saving the Map with the errors stops the Map from being run because it can’t compile until the errors are corrected.

 

Issue/Introduction

TIBCO Scribe® Online is adding support for Unicode 3.0 in the Summer 2017 release of TIBCO Scribe® Online. If you have letters from other languages, such as Cyrillic characters, in Entity or Field Names in the datastore or you are using Unicode characters from a version of Unicode higher than 3.0, you may experience Map validation errors.

Resolution

Resolution


There are two options for solving this issue:


Option 1:
  1. Create a new Connection to the datastore that has entities or fields with Unicode characters.
  2. Reassign the old Connection to the new Connection in all Maps.
  3. Edit the Maps to fix the links to entities and field mappings to use the new names.

NOTE: If you are using Unicode 4.0 or higher characters, Option 1 will not fix the issues.  Either do not edit Maps after resetting metadata or use Option 2.


Option 2:


Edit the Scribe.Core.ProcessorService.exe.config file for each Agent to use the old method of managing special characters, as follows:
 
  1. Stop the Agent service.
  2. Navigate to the On-Premise Agent folder, typically C:\Program Files (x86)\Scribe Software\Scribe Online Agent
  3. Locate the Scribe.Core.ProcessorService.exe.config file and open it in a text editor, such as NotePad++.
NOTE: Make sure to run the text editor as Administrator or the changes to the file may not be saved.
  1. Scroll down to the <appSettings> section and add the following:
<add key="Scrub" value="[^a-zA-Z0-9_]" />

NOTE: Different values can be used, but check with TIBCO Scribe first.
 
kA332000000U3GRCA0_en_US_1_0
 
  1. Save the changes to the file. These changes are honored when the Agent is upgraded in the future.
  2. Restart the Agent service.
  3. Reset metadata again so that the prior method of working with special characters sets the entity and field names back to the format expected by the Maps.
  4. Open one or more of the affected Maps and click Validate to make sure there are no errors.
  5. Run the Solution again.