Long time no see!

So I'm still in education metadata (and there to stay) and my interest in Linked Data approaches to solve challenges in that field hasn't faded a bit. Recently I was examining the Europass Learning Model. As happy as I was that it is an extension of the W3C Verifiable Credentials Data Model (so no total reinvention) so sad it is to see that there are no answers to open issues and reactions to Pull Requests in their GitHub Repository.

Those of you who clicked the above link to the Europass Learning Model might have noticed that some links are giving 404 errors. That's something which can always happen. But it should maybe have been fixed for a model released over two years ago by the European Commission (maybe it wouldn't have happened if there would be an engaging community maintaining it).

But that aside. I wanted to see what kind of vocabularies are used in the model. Since I did not find any information about this somewhere (might have missed it), I thought I just parse it out of the examples (the examples and XSD-files seem a better source of the model than the documentation, but that's for another post).

So let's do it!

Let's have a look at the raw example file.

The vocabulary is always given by a targetFrameworkUrl Attribute, e.g.

<type targetFrameworkUrl="http://data.europa.eu/snb/credential/25831c2" uri="http://data.europa.eu/snb/credential/6dff8a0f87">

So the idea is quite simple: Parsing the whole XML with XPath for the targetFrameworkUrl attributes. We will use Javascript, so we can do it directly in the browser:

We can use the DOMParser interface to accomplish this.

var parser = new DOMParser()

First we need to get the text content out of the <pre> HTML tag:

var txt = document.querySelector('pre')

Now we parse the document:

var xmlDoc = parser.parseFromString(txt, 'text/xml')

Now we should be able to parse the document:

var res = xmlDoc.evaluate('//@targetFrameworkUrl', xmlDoc')

Let's check what we got with the iterateNext() method of the XPathResult object

res.iterateNext()
//  targetFrameworkUrl="http://data.europa.eu/snb/credential/25831c2"
res.iterateNext()
// targetFrameworkUrl="http://publications.europa.eu/resource/authority/country"

Looks like what we want. To make more use of the result, we change the result type of the XPathResult to ORDERED_NODE_SNAPSHOT_TYPE. This way it is easier to navigate through the results. For the different result types, see here:

var res = xmlDoc.evaluate('//@targetFrameworkUrl', xmlDoc, null, XPathResult.ORDERED_NODE_SNAPSHOT_TYPE)

res

// XPathResult { resultType: 7, invalidIteratorState: false, snapshotLength: 236 }

Now we already see a length attribute there. Nice!
Lets parse the results so we get a nice list containing no duplicates:

var vocabsUrls = new Set()

for (var i = 0; i < res.snapshotLength; i++) {
    vocabsUrls.add(res.snapshotItem(i).nodeValue)
}

This gives us a nice list of the used vocabularies:

Array.from(vocabsUrls).join('\n')
// Output:
"http://data.europa.eu/snb/credential/25831c2
http://publications.europa.eu/resource/authority/country
http://publications.europa.eu/resource/authority/human-sex
http://data.europa.eu/snb/supervision-verification/25831c2
http://data.europa.eu/snb/learning-opportunity/25831c2
http://data.europa.eu/snb/isced-f/25831c2
http://publications.europa.eu/resource/authority/language
http://data.europa.eu/snb/learning-assessment/25831c2
http://data.europa.eu/snb/learning-setting/25831c2
http://data.europa.eu/snb/eqf/25831c2
http://data.europa.eu/snb/qdr/25831c2
http://data.europa.eu/snb/skill-type/25831c2
http://data.europa.eu/snb/skill-reuse-level/25831c2
http://external.skill-framework.eu
http://data.europa.eu/esco/model#Skill
http://data.europa.eu/snb/learning-activity/25831c2
http://data.europa.eu/snb/assessment/25831c2
http://data.europa.eu/snb/entitlement/25831c2
http://data.europa.eu/snb/entitlement-status/25831c2
http://publications.europa.eu/resource/authority/file-type
http://data.europa.eu/snb/encoding/25831c2"