IBM® Integration Bus uses
the DataObject parser to read and write messages that belong to the
DataObject domain.
You
must use the DataObject domain when you use the WebSphere® Adapter nodes
in your message flow. IBM Integration Bus uses the
DataObject parser to read and write messages, or business objects,
from Enterprise Information Systems (EIS) such as SAP. The DataObject
parser uses either XML schema files (.xsd) from an IBM Integration Bus library,
or message definitions in a message set when reading and writing messages.
A message definition (.mxsd) file is an annotated XML schema file.
The TX Map node
can consume and create messages in the DataObject domain.
To
create a message flow that uses the TX Map node
to create or consume messages in the DataObject domain, do the following
steps within the IBM Integration Toolkit:
-
In the IBM Integration Development perspective, use the
Adapter Connection wizard to model the business object as either:
- A global element in an XML schema file (.xsd) within a broker library
- A message in a message definition (.mxsd) file, within a broker message set
For detailed instructions, see the IBM Integration Bus product
documentation.
- In the Transformation Extender Development
perspective,
use Map Designer to
create a map, and add input or output cards as required. In the New
Card dialog, select the message definition (.mxsd) file
or the XML schema (.xsd) file as the Type tree,
and the group corresponding to the business object message as the Type.
- Switch to the IBM Integration Development perspective.
Create a message flow that contains a TX Map node
that uses your map.
- If the message is consumed by an input card, configure
the input node of the message flow to specify the Message
Domain as DataObject.
- If the message is created by an output card, configure
the corresponding TX Map node
output terminal to specify the Message Domain as DataObject.
Set the Encoding and Coded Char
Set ID properties to match the byte order (endianness)
and the character set of the output data from the map.
- Complete the message flow and deploy it to the broker,
along with the message set.
If the message is to be consumed by an input card, at run
time the input node of the message flow starts the DataObject parser.
The DataObject parser uses the XML schema or the message set to create
a message tree from the EIS business object. The
TX Map node
receives the DataObject domain message tree and passes it to the input
card. The map uses the tree directly when it runs.
If the message
is to be created by an output card, at run time the map creates the
DataObject domain message tree directly. The TX Map node
propagates the message tree to the next node in the message flow.
The output node of the message flow starts the DataObject parser,
which uses the XML schema or the message set to create the EIS business
object.
You do not have to create an equivalent type tree for
the business object message. The map is built with the message definition
file. At run time, the map operates directly on the DataObject domain
message tree, and not on a bit stream.