changing default encoding of the software

Document ID : KB000112113
Last Modified Date : 02/10/2018
Show Technical Document Details
Issue:
I wanted to see if we can replace our default enconding on harvest to UTF-8. we are seeing some weird charcters when CA Release automation generates reports from harvest server which is also causing some of our workflows to not work properly.
Environment:
CA Harvest SCM Client v12.x and up
Windows OS
Cause:
The problem turned out to be non-ASCII characters in package names.  When a package is first created, the default package name is assigned.  But later, during development, the package name is changed to reflect the contents of the package.  The users will copy and paste from Excel, Word, or other applications when they change the package name, and this will sometimes result in the non-ASCII characters they are seeing.  SCM handles this correctly, and the package name is stored and displayed with no problems.  However, when another application such as Release Automation queries the Harvest database and retrieves these package names, it sometimes handles these characters incorrectly.  This is the problem to be solved.

The first proposed solution was to add the following line to the %CA_SCM_HOME%\workbench.ini:

-Dfile.encoding=utf-8

However this did not resolve the problem.
Resolution:
The solution that was chosen was to create a custom script that would examine the names of selected packages and return an error code if any package names were found to have non-ASCII characters.  This could be set up as a pre-link UDP to the Promote process, thus preventing any packages with problematic package names from being promoted to the Production state until the package name was corrected.