Error indicates "An invalid XML character" when handling data that originated from a WebSphere Adapter
Unable to process data from source containing unicode data due to a serialization error in broker indicating "An invalid XML character (Unicode: 0x0) was found in the element content"
When processing data using your broker that originated from a WebSphere JCA Adapter, you encounter an exception of the following form:
Source = java.lang.RuntimeException caught trying to serialize BO BusinessObject: F4211BeginDoc@382a382a (F4211FSBeginDoc=F4211FSBeginDoc@67e667e6) probeid = 214 Stack Dump = java.lang.RuntimeException: An invalid XML character (Unicode: 0x0) was found in the element content:
This error occurs while doing data transformations within your broker, either for an XSLT or to enable monitoring or tracing of the data stream, and results in a flow failure.
The problem occurs when the source data contains unicode data which is not allowed in XML. In the above example, this is the unicode character 0x0, but there are a number of other unicode characters that can cause the same result. The problem does not occur in the WebSphere Adapter itself because the WebSphere Adapters are entirely capable of handling full unicode data. Rather, the problem happens in some brokers which fail to properly serialize the incoming data into a valid XML object.
The following tech note relates to this problem: