def dbpedia_sparql_extract_people(people_list_file):
# With help from https://rdflib.github.io/sparqlwrapper/
# and https://stackoverflow.com/questions/38332857/
# sparql-query-to-get-all-person-available-in-dbpedia-is-showing-only-some-person
if os.path.exists(people_list_file):
os.unlink(people_list_file)
# total_people = dbpedia_sparql_get_people_count()
total_people = 2109301
for i in range(0, total_people, 10000):
people_list = []
offset = str(i)
print("We're at {sofar} out of {total}".format(sofar=offset, total=total_people))
sparql_query = """
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX dbo: <http://dbpedia.org/ontology/>
PREFIX dbp: <http://dbpedia.org/property/>
SELECT ?name
WHERE { ?resource foaf:name ?name .
?resource rdf:type dbo:Person .
}
"""
sparql_query_offset = "LIMIT 10000 OFFSET {}".format(offset)
response = util.dbpedia_do_sparql_query(sparql_query + sparql_query_offset)
results = response['results']['bindings']
people_list.extend([res['name']['value'] for res in results])
print("Adding {count} to people list file".format(count=len(results)))
with open(people_list_file, 'a') as f:
f.writelines("\n".join(people_list))
def dbpedia_sparql_get_people_count() -> int:
sparql_query = """
PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
PREFIX dbo: <http://dbpedia.org/ontology/>
PREFIX dbp: <http://dbpedia.org/property/>
SELECT COUNT(*)
WHERE { ?resource foaf:name ?name .
?resource rdf:type dbo:Person .
}
"""
res = util.dbpedia_do_sparql_query(sparql_query)
return int(res['results']['bindings'][0]['callret-0']['value'])