The connector supports the exchange of documents in XML, EDI, and binary
formats. The TPI connector uses data handlers to convert TPI documents
to WebSphere Business Integration Adapter business objects, and to convert
WebSphere Business Integration Adapter objects to TPI-supported document
formats.
Table 4. WebSphere Business Integration Adapter-delivered data handlers
Document format | WebSphere Business Integration Adapter-delivered data handler |
---|---|
XML | XML |
EDI | EDI |
Binary | Any |
The connector calls the appropriate data handler by passing in the document MIME type to the DataHandler class. Additionally, the connector may pass the BOPrefix value, if it is populated in the incoming business object. When processing an inbound document, the connector gets the MIME type of the document from the trading partner configuration file. When processing an outbound request business object, the connector gets the MIME type from the trading partner configuration file, based on the values of the DocumentType and ReceiverID attributes in the child meta-object. See Business object structure for more information.
Because the TPI connector calls data handlers to convert from streams to business objects and from business objects to streams, each business object must conform to the specifications of the data handler called to perform the conversion. For information about the specific business object requirements for a particular data handler, see the Data Handler Guide.
The TPI Server has specific requirements for the content and formatting of header information in EDI and XML documents. All XML and EDI documents must contain a valid SenderID, RecieverID and a unique CycloneID. If the SenderID or ReceiverID values are invalid, or if the CycloneID value is not unique in the TPI system, the TPI Server will not process the document. These values must correspond to request business objects so that the data handler places them correctly in the document.
For EDI documents the placement of these values is mandated by the EDI specification. Consider the following sample EDI header:
ISA*00* *01*XXXXXX *L1*2 *L0*0*961106*2106*U*00302*000087875*
The SenderID is read from two columns, *L1*2 in this example. The first column must contain the two character SenderID qualifier. The second column must contain all remaining characters of the SenderID.
Like SenderID, the ReceiverID is also read from two columns in the EDI document header, *L0*0 in this example. The first column must contain the two character ReceiverID qualifier. The second column must contain all remaining characters of the ReceiverID.
for EDI documents, the CycloneID corresponds to the EDI Control ID. The CycloneID is represented in a single column, *000087875* in this example. This value corresponds to the EDI Control ID number 87875.
The placement of SenderID, ReceiverID, and CycloneID in XML documents can be customized using the TPI Server Administrator. See the Administrator's Guide included with the TPI Server for more information.