SAP CPI: How to read MIME Multipart response message received from Ariba Network

[YayatiEkbote]

Hello Experts,

I am working on fetching Contract Request cXML from Ariba Network. I need to convert the cXML into JSON to send to target system. But the message which I am getting from Ariba Network is of this format (I have removed/altered some actual content) -

------=_Part_8869_950514950.1642859653328
Content-Type: text/xml; charset=UTF-8
Content-ID: XXXXXXXXXX.XXXXXXXX@app1008.snv.ariba.com

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE cXML SYSTEM "http://xml.cxml.org/schemas/cXML/1.2.052/Private.dtd">
<cXML timestamp="2022-01-22T05:54:13-08:00" payloadID="9999999999999-6666666666666835933@77.777.77.777">
	<Response>
		<Status code="200" text="OK"/>
		<DataResponse>
			<Attachment>
				<URL>XXXXXXXXXX.XXXXXXXX@app1008.snv.ariba.com</URL>
			</Attachment>
		</DataResponse>
	</Response>
</cXML>

------=_Part_8869_950514950.1642859653328
Content-Type: text/xml; charset=UTF-8
Content-ID: XXXXXXXXXX.XXXXXXXX@app1008.snv.ariba.com

<?xml version="1.0" encoding="UTF-8"?>
<!DOCTYPE cXML SYSTEM "http://xml.cxml.org/schemas/cXML/1.2.042/Contract.dtd">
<cXML payloadID="1572609866292.423681963.000001843@NMC8pBEXStATEUCcgVjhR3bzQ00=" timestamp="2019-11-01T05:04:26-07:00" version="1.2.042" xml:lang="en-US">
	<Header>
	</Header>
	<Request deploymentMode="test">
		<ContractRequest>
			<ContractRequestHeader>
			</ContractRequestHeader>
			<ContractItemIn operation="new">
			</ContractItemIn>
		</ContractRequest>
	</Request>
</cXML>
------=_Part_8869_950514950.1642859653328--

How to read the actual cXML content from this entire Multipart message? How to parse it? Which message transformer should be used?

I tried using MIME Multipart decoder but it is not doing anything.

I tried setting the flag and removing too - Multipart Headers Inline but it doesn't change anything.

In headers, I am getting the boundary of Multipart message in header. In groovy script, I can split the message at boundary but after that still the Content-Type and Content-ID before prolog of cXML are present in the split message. As the content-ID will be unique in every message, I am unable to find a decent/standard way of reading cXML part and pass it to message mapping step.

Regards,

Yayati Ekbote

[daviddasilva]

Hi,

If possible, I would use a splitter so that you can process each "part" individually. Then I would store the Content-type and Content-ID as custom headers, followed by a filter/groovy script to get just the XML section where I would then do whatever needs to be done to the XML part such as a message mapping if required. Then use the standard XML to JSON transformer. You can then send these messages one by one to the destination.

Kind regards,

David