Reading an XML file into an Interface

Question

Hi I have a need to be able to select an XML file, and use fields in it to populate an interface. 
 At the moment I have a simpe interface with a File Upload and a Rule Input of xmlData (Document) which the file upload saves to. 
 I am then kicking off a process model on submit which runs a script task. 
 I have activity class parameters for the fields I want and am running the following script on each one

=xpathdocument(pv!xmlFile,"/Document/Fields/Field[@Name='Department']/@Value") 
 
 Currently I just get an error on the process model when I debug:

Script Task 
 An error occurred while evaluating expression: =xpathdocument(pv!xmlFile,"/Document/Fields/Field[@Name='Department']/@Value") (Expression evaluation error at function 'xpathdocument': An error occurred while trying to parse the XML stream.) (Data Inputs)

Does anyone know what the error is more specifically? Am I on the right lines or is there a better way of doing this?

Acacio Barrado · Accepted Answer

Hi David,

The xml example that you provided is not a valid xml, the main problem is that you cannot have a XML tag name with spaces, otherwise the second word would be interpreted as an attribute.

I edited your example in order to make the XML valid, and from there you can adjust to your process. Please find the valid xml in the code below:

<?xml version="1.0" encoding="UTF-8"?>
<Root>
	<OriginalVariables>
		<CardholderElement>Carolina Kenny</CardholderElement>
		<CardNumberElement>449450294</CardNumberElement>
	</OriginalVariables>
	<Scanning/>
	<QueueInfo>
		<Company>Parkmill</Company>
	</QueueInfo>
</Root>

Below is the code example for you get the company value:

xpathdocument(
  docId: YOURDOC,
  expression: "Root/QueueInfo/Company/text()" 
)

In case these responses were helpful don't forget to mark the question as answered as it can also help other that are facing similar issues.

Regards,

Acacio B.

Mike Schmitt · Answer

David Craven said: How would I find the DocumentID? 
 Your assumptions are correct, as long as you account for needing to submit a form prior to accessing a document's properties (which i believe you're already doing correctly). If needed, you can always typecast a document object directly to its ID by, among other things, passing the document object through the "toInteger()" function. As I noted in my prior reply though, in almost all cases it won't make any difference, as integer and document objects will almost seamlessly cross-typecast as needed depending on the service in question.

Acacio Barrado · Answer

Hi David, 
 Yes, there is a big difference in relation to the format of your XML. 
 In the current format as you sent in your example, the data that you are trying to retrieve is stored in one of the attributes from of the field called Field. This field is generated multiple times in your XML and the only difference between them is the values in the attributes (Field Name and value). It is possible to get the value that you are looking for, by AFAIK, you would need to rely on indexes, I&rsquo;m attaching a code below where that you can use to get the Company value: 
 xpathdocument(
 docId: YOURDOC,
 expression: "Document/Fields/Field[2]/@Value" 
) 
 
 On the other hand, using the second format that you proposed you could easily get the value from company without need to rely on index. I&rsquo;m attaching a code example below: 
 
 xpathdocument(
 docId: YOURDOC,
 expression: "Document/Fields/Company/text()" 
) 
 
 Regards 
 
 Acacio B.

Reading an XML file into an Interface

Top Replies