def show(self, id, format="html"):
"""GET /vocabulary/search/id: Show a specific item"""
# url('vocabulary_search', id=ID)
import re
id = re.sub("http:/", "http://", id)
query = """
SELECT DISTINCT rdfdoc.id, rdfdoc.uri
FROM rdfdoc, stat_result, rdf_property_stat, rdf_property
WHERE rdf_property.uri='%s' AND
rdf_property.id=rdf_property_stat.rdf_property_id AND
rdf_property_stat.stat_result_id=stat_result.id AND
stat_result.rdfdoc_id=rdfdoc.id
ORDER BY rdfdoc.id;
""" % (
id
)
try:
datasets = []
result = Session.execute(query)
for row in result:
obj = {"id": row[0], "uri": row[1]}
datasets.append(obj)
return json.dumps(datasets)
except:
return json.dumps("")
def _rankSuggestionLodstats(self, suggestionUri, entities):
#The most time consuming - implement caching here
#just cache to /tmp for now
onlyclasses = set()
for entity in entities:
onlyclasses.add(entity['class'])
onlyclasses = list(onlyclasses)
cacheId = uuid.uuid5(uuid.NAMESPACE_URL, suggestionUri.join(sorted(onlyclasses)).encode('utf-8'))
cachePath = '/tmp/'
cacheNamespace = 'suggestionsCache'
cacheEntry = str(cachePath) + str(cacheNamespace) + str(cacheId)
if(os.path.exists(cacheEntry)):
return pickle.load(open(cacheEntry, 'rb'))
propertyQuery = """SELECT stat_result_id
FROM rdf_property_stat, rdf_property
WHERE rdf_property.id=rdf_property_stat.rdf_property_id
AND rdf_property.uri='%s';""" % suggestionUri
q = Session.execute(propertyQuery)
propertyDatasets = set()
for row in q:
propertyDatasets.add(row[0])
entitiesDatasets = set()
for entityUrl in onlyclasses:
classQuery = """SELECT stat_result_id
FROM rdf_class_stat_result, rdf_class
WHERE rdf_class_stat_result.rdf_class_id=rdf_class.id
AND rdf_class.uri='%s';""" % entityUrl
q = Session.execute(classQuery)
for row in q:
entitiesDatasets.add(row[0])
propertyQuery = """SELECT stat_result_id
FROM rdf_property_stat, rdf_property
WHERE rdf_property.id=rdf_property_stat.rdf_property_id
AND rdf_property.uri='%s';""" % entityUrl
q = Session.execute(propertyQuery)
for row in q:
entitiesDatasets.add(row[0])
common = propertyDatasets.intersection(entitiesDatasets)
pickle.dump(len(common), open(cacheEntry, 'wb'), protocol=pickle.HIGHEST_PROTOCOL)
return len(common)
