I’m trying to execute an Ultra pipeline with an XML payload, but I’m getting the error below:
Document input views only accept JSON-encoded data
Resolution
Please raise a defect with the following information, Pipeline Ruuid:d3c6dcc2-a498-4517-8df9-54910d504959
Reason
Snap failed unexpectedly and did not provide any reason
What steps can be taken to make Ultra parse an XML payload?
You can place an XML Parser snap at the head of the pipeline to parse the XML request content. To send an XML response back to the client, you’ll need to use an XML Formatter with an unlinked output and clear the ‘Root element’ field so that a separate binary document will be created for every document that flows through the XML Formatter.
All my existing pipelines (which were functional in non-Ultra) were using XML Parser snaps already… My client program which was making requests to SnapLogic was setting the “Content-Type” header to “application/xml” which I suspect might have been causing this error.
I switched my client program to send a JSON payload and also modified the Content-Type header to send application/json and my Ultra pipeline accepts the payload.
Can you confirm if an Ultra pipeline that is sent an XML payload should also set the Content-Type header to application/xml or will that cause problems?
The Content-Type should always match what is in the payload.
@tstack, thanks so much for your input… so I’ve built the following pipeline which successfully executes in Ultra:
In my original pipeline I was passing in a URL parameter which was being captured as a pipeline parameter. In Ultra, I want to capture this pipeline parameter by passing a parameter to the “Pipeline Execute” snap in my pipeline. However, based on the documentation for Ultra… I have two questions:
I think you can, but it’s not as easy as it oughta be right now. These values are available in the header of the binary document that is sent into the pipeline. Unfortunately, the XML Parser doesn’t give you access to the header. Instead, you can use a ‘Binary To Document’ snap to convert the binary document and its header into a regular document. By default, the body of the request is converted into a base64 string and placed in the ‘content’ field. The header data is added directly to the document. As an example, an ultra pipeline with a ‘Binary To Document’ snap and a Mapper that maps ‘$’ to ‘$content’ returns the following when a request is sent:
curl -k --data-binary '<foo><bar>hi</bar></foo>' -H 'Content-Type: text/xml' 'https://feed1:8082/api/1/rest/feed-master/queue/snaplogic/projects/shared/ts-xml-ultra-task?param1=value1' | jq .
{
"task_name": "snaplogic/projects/shared/ts-xml-ultra-task",
"content-length": "24",
"method": "POST",
"query": {
"param1": [
"value1"
]
},
"message_id": "8e22c67515a888254fdb859c090c2c6124bfafca-22392@feed1",
"uri": "https://feed1:8082/api/1/rest/feed-master/queue/snaplogic/projects/shared/ts-xml-ultra-task?param1=value1",
"content": "PGZvbz48YmFyPmhpPC9iYXI+PC9mb28+",
"accept": "*/*",
"client_port": 59455,
"path_info": "",
"host": "feed1:8082",
"server_ip": "10.1.10.222",
"content-type": "text/xml",
"client_ip": "10.1.10.222",
"server_port": 8082,
"user-agent": "curl/7.45.0"
}So, you can send the output of the ‘Binary To Document’ into the PipeExec, passing whatever values you want as pipeline parameters. In the child pipeline, it should start with a Document To Binary snap to get back the original XML stream of data and then an XML Parser to parse that.
Sorry, it’s a bit roundabout. The focus has mostly been on requests being JSON-based.
