Hi,

Check below two links regarding JVM arg -Dfile.encoding=UTF8. These may help you.

https://stackoverflow.com/questions/16888616/change-default-charset-of-jvm-to-utf-8

https://stackoverflow.com/questions/6031877/jvm-property-dfile-encoding-utf8-or-utf-8

Changing the JVM command line argument "file.encoding" may help you with your current issue, but as it has impact on the environment as a whole, it may break things elsewhere.

You should check how the CSV is written to file and try to set the desired target encoding there. Perhaps you should post the complete code sequence so we can see what you are attempting.

Just out interest: what characters do you find are missing in UTF-8 ? And what encoding are you planning on using instead ?

My understanding is that UTF-8 is universal ?

See here for instance: http://https://superuser.com/questions/946612/what-languages-does-the-character-encoding-utf-8-support

In any case:

I took a look at the Activity "@baseclass.pxConvertResultsToCSV|Pega-RulesEngine:07-10-31" and I believe the reason the file is being written with the default encoding is due to this call (in step 5).

 PRWriter out = new PRWriter(stream, null, false);

The 'null' fits into the slot for encoding; I think maybe it is falling back on the system encoding, if you don't provide a specific encoding.

The Activity itself is 'Final' - so you cannot save a copy with the same name.

You can save it under a different name (say "myConvertResultsToCSV") in your own ruleset.

Try this: and change the line in step 5 to replace the null with whichever encoding you need here.

Please check the Javadocs though - they mention that the encoding value should be one of the type listed in "PRFileUtil.ENCODING_*".... which is quite a small list.