Example to grab predicate counts from a sparql endpoint

gistfile1.md

Since @mjg shamed me into posting a gist of this, here's a simple example of something gross.

Query dbpedia endpoint for most highly used properties, get the jq response format, grab the counts and values, format them usefully in sed, filter out the top 50 in awk, and print to the screen:

    curl -H 'Accept: application/sparql-results+json' -d 'query=SELECT+?p+(COUNT+(?p)+AS+?pCount)+WHERE+{+?s+?p+?o+}+GROUP+BY+?p+ORDER+BY+DESC(?pCount)' http://dbpedia.org/sparql? | jq '.results.bindings[] | [.p.value, "|", .pCount.value] | add' | sed 's/\"//g' | awk ' BEGIN { FS = "|" ; OFS = "|" } { if ($2 >= 50) { print $0 } } ' | less

👏

This could also be done without jq or the awk filtering by requesting CSV and using a HAVING clause:

curl -H 'Accept: text/csv' -d 'query=SELECT+(CONCAT(STR(?p),"|",COUNT(?p))+AS+?v)+WHERE+{+?s+?p+?o+}+GROUP+BY+?p+HAVING+(COUNT(?p)+>+50)+ORDER+BY+DESC(COUNT(?p))' 'http://dbpedia.org/sparql?' | sed 's/\"//g'