How to capture piece of XML to a literal as is? about sparql.anything

[VladimirAlexiev]

@enridaga Yes!

Note: It's not the whatwg namespace, but the xhtml namespace.
Actually I've renamed them in my script, see the comments below as to where they appear:

prefix x:     <http://www.w3.org/1999/xhtml#>              # raw data prefix in HTML->RDF, prepended to any other prefix, eg x:gml:featurecollection, x:plu:indicationreference, x:nilreason, x:xsi:nil
prefix html:  <https://html.spec.whatwg.org/#>             # html:innerHTML, html:innerText

[enridaga]

I mean the complete XML structure with sub-elements, not just its text content

Interesting use case, unfortunately, this can't be done with the XML triplifier. Maybe you can try to pretend it's HTML, I think that triplifier should have an innerHTML property that in principle should return the GML code. See the documentation for details and let us know if it solves your problem.

[VladimirAlexiev]

@enridaga It kind of works, but not quite. See test files at https://gist.github.com/VladimirAlexiev/318ab4924c756ead618ebdded6428509 and compare

• sparql-anything-test-xml.ttl
• sparql-anything-test-html.ttl

I use these prefixes:

prefix x:     <http://www.w3.org/1999/xhtml#>              # raw data prefix in HTML->RDF, prepended to any other prefix, eg x:gml:featurecollection, x:plu:indicationreference, x:nilreason, x:xsi:nil
prefix html:  <https://html.spec.whatwg.org/#>             # html:innerHTML, html:innerText
prefix xlink: <http://www.w3.org/1999/xlink#>
prefix gml:   <http://www.opengis.net/gml/3.2#>
prefix plu:   <http://inspire.ec.europa.eu/schemas/plu/4.0#>
prefix xyz:   <http://sparql.xyz/facade-x/data/>           # raw prefix in XML->RDF data
prefix fx:    <http://sparql.xyz/facade-x/ns/>             # fx:properties in SERVICE call, fx:root in output
prefix rdf:   <http://www.w3.org/1999/02/22-rdf-syntax-ns#># each XML element becomes rdf:type
prefix rdfs:  <http://www.w3.org/2000/01/rdf-schema#>      # rdfs:member
prefix xsd:   <http://www.w3.org/2001/XMLSchema#>
prefix xsi:   <http://www.w3.org/2001/XMLSchema-instance#> # xsi:nil "true": filter out such elements. xsi:schemaLocation

Problems:

• all elements are lowercased: both inside innerHTML and in RDF
• all elements in RDF are prepended the x: prefix, eg x:gml:featurecollection, x:plu:indicationreference, x:xsi:nil
  (the originals are gml:FeatureCollection, plu:indicationReference, xsi:nil`

I guess I can live with this, but the script will be completely non-portable between XML and HTML input...

[enridaga]

I guess I can live with this, but the script will be completely non-portable between XML and HTML input...

Of course, this is a hack, XML is different from HTML.

We can open issues to support the case better, when using the HTML triplifier with XML content. To summarise, these seems to be the problems:

[ ] xmlns is not considered a ns prefix, and the default whatwg namespace is prepended
[ ] element names are lowercased, while in XML tags are case sensitive

[enridaga]

@enridaga Yes!

Note: It's not the whatwg namespace, but the xhtml namespace. Actually I've renamed them in my script, see the comments below as to where they appear:

prefix x:     <http://www.w3.org/1999/xhtml#>              # raw data prefix in HTML->RDF, prepended to any other prefix, eg x:gml:featurecollection, x:plu:indicationreference, x:nilreason, x:xsi:nil
prefix html:  <https://html.spec.whatwg.org/#>             # html:innerHTML, html:innerText

My suspicion is that this depends on the fact that the HTML parser (beatifulsoup) does not handle prefixes. Will investigate.