Open danielmlow opened 3 years ago
pyld
would the tool to use. there is a function in reproschema-py
that will do the "right thing" to reading any jsonld file that is on the filesystem (i.e. data served locally). and rdflib can definitely be used to query the data or convert to a different form.
If we just use the
json
package, we get something like this for a single activity (covid19 questionnaire) of a single submission:So then to build the dataset, I'd take every
@type
section that is equal toreproschema:Response
, then obtain the item name fromisAbout
and get the values, we'd get something like this:(1 row=1 protocol)
Other things that could be added are timestamps for each item/activity, language (which should be the same, but maybe they switch half way)
@satra, would it be more useful to parse use something like rdflib-jsonld to obtain the graph?