Search code examples
oracle-databaseunicodeencodingadapterbpel

Inserting UTF-8 characters into database via BPEL process

We have a BPEL process that is taking characters (e.g. Chinese) and attempts to insert them into an NVARCHAR column. The XML payload looks correct throughout the various stages (and we see the encoding is UTF-8) but when it comes to the actual INSERT we are seeing incorrect characters. Our theory is that the DB Adapter could be the problem but we do not know for sure.

Is there a way to examine the version and config of the DB Adapter, to see if it can handle UTF-8 and Unicode? We have successfully inserted Chinese and Hewbrew characters into the database using TOAD, Java servlets, etc. so the database itself is configured. It is Oracle version 9.2.0.8.0.

It may be that the fault is elsewhere than the DB Adapter so we are trying to eliminate things. Would it be worth trying to look at the HTTP header to check the encoding here?

Thanks

Solution

  • We got to the bottom of this in the end. In a WSDL file (as part of the database connection details) we specified the defaultNChar=true property as an attribute and this was passed on to the database driver.