curl example on ifcOWL page fails when copied + suggestion for RDF access via content negotiation

[kathrinrin]

Great that bSDD supports RDF!

I noticed a small issue while testing the curl command from https://technical.buildingsmart.org/standards/ifc/ifc-formats/ifcowl/:

curl -X 'GET' 
  'https://api.bsdd.buildingsmart.org/api/Class/v1?Uri=https%3A%2F%2Fidentifier.buildingsmart.org%2Furi%2Fbuildingsmart%2Fifc%2F4.3%2Fclass%2FIfcWall' 
  -H 'accept: text/turtle'

Copy/pasting this command into a shell does not work because the line breaks are not escaped. This version works:

curl -X 'GET' \
  'https://api.bsdd.buildingsmart.org/api/Class/v1?Uri=https%3A%2F%2Fidentifier.buildingsmart.org%2Furi%2Fbuildingsmart%2Fifc%2F4.3%2Fclass%2FIfcWall' \
  -H 'accept: text/turtle'

Another simplification would be to let curl handle the URL encoding:

curl -G \
  --data-urlencode "Uri=https://identifier.buildingsmart.org/uri/buildingsmart/ifc/4.3/class/IfcWall" \
  -H "accept: text/turtle" \
  https://api.bsdd.buildingsmart.org/api/Class/v1

Finally, according to Linked Data principles, an even simpler pattern would be if the identifier URI supported content negotiation (similar to how it already returns text/html), e.g.:

curl -H "Accept: text/turtle" \
  https://identifier.buildingsmart.org/uri/buildingsmart/ifc/4.3/class/IfcWall

If the identifier URI supported RDF representations via content negotiation, the RDF would be much easier to consume for Linked Data applications.

[atomczak]

Thanks @kathrinrin.

True, the escaping is lost when converting the source markdowns to our website, so as a workaround, I turned it into a single-line command. I also followed your advice to let curl handle the URL encoding. See: https://github.com/buildingSMART/technical.buildingsmart.org/blob/main/ifcOWL.md#bsdd-supports-the-rdfttl

About content negotiation, that is true, but this will take a bit longer time to implement - added to our backlog.