def _nq_row(triple, context):
if isinstance(triple[2], Literal):
return "%s %s %s %s .\n" % (triple[0].n3(), triple[1].n3(),
_quoteLiteral(triple[2]), context.n3())
else:
return "%s %s %s %s .\n" % (triple[0].n3(), triple[1].n3(),
triple[2].n3(), context.n3())
def _nq_row(triple, context):
if isinstance(triple[2], Literal):
return u"%s %s %s %s .\n" % (
triple[0].n3(), triple[1].n3(),
_xmlcharref_encode(_quoteLiteral(triple[2])), context.n3())
else:
return u"%s %s %s %s .\n" % (triple[0].n3(), triple[1].n3(),
_xmlcharref_encode(
triple[2].n3()), context.n3())
def _nq_row(triple,context):
if isinstance(triple[2], Literal):
return u"%s %s %s %s .\n" % (triple[0].n3(),
triple[1].n3(),
_xmlcharref_encode(_quoteLiteral(triple[2])),
context.n3())
else:
return u"%s %s %s %s .\n" % (triple[0].n3(),
triple[1].n3(),
_xmlcharref_encode(triple[2].n3()),
context.n3())
def n3(node: Union[URIRef, BNode, Literal]) -> str:
"""Method takes an RDFLib node of type BNode, URIRef, or Literal and serializes it to meet the RDF 1.1 NTriples
format.
Src: https://github.com/RDFLib/rdflib/blob/c11f7b503b50b7c3cdeec0f36261fa09b0615380/rdflib/plugins/serializers/nt.py
Args:
node: An RDFLib
Returns:
serialized_node: A string containing the serialized
"""
if isinstance(node, Literal): serialized_node = "%s" % _quoteLiteral(node)
else: serialized_node = "%s" % node.n3()
return serialized_node
def serializeQuad(g):
"""
replacement for graph.serialize(format='nquads')
Still broken in rdflib 4.2.2: graph.serialize(format='nquads')
returns empty string for my graph in
TestGraphFromQuads.testSerializes.
"""
out = []
for s, p, o, c in g.quads((None, None, None)):
if isinstance(c, Graph):
# still not sure why this is Graph sometimes,
# already URIRef other times
c = c.identifier
if '[' in c.n3():
raise ValueError(str(c))
ntObject = _quoteLiteral(o) if isinstance(o, Literal) else o.n3()
out.append("%s %s %s %s .\n" % (s.n3(), p.n3(), ntObject, c.n3()))
return ''.join(out)
def _statement_encode(self, xxx_todo_changeme, context):
"""helper function to encode triples to sesame statement uri's"""
(s, p, o) = xxx_todo_changeme
query = {}
url = self.url + '/statements'
if s:
query['subj'] = s.n3().encode('utf8')
if p:
query['pred'] = p.n3()
if o:
query['obj'] = _quoteLiteral(o.n3())
# o.n3()
# quote_plus(o.n3().encode("utf-8"))
if context:
# TODO FIXME what about bnodes like _:adf23123
query['context'] = "<%s>" % context
if query:
url = url + "?" + urlencode(query)
return url
def _statement_encode(self, xxx_todo_changeme, context):
"""helper function to encode triples to sesame statement uri's"""
(s, p, o) = xxx_todo_changeme
query = {}
url = self.url + '/statements'
if s:
query['subj'] = s.n3().encode('utf8')
if p:
query['pred'] = p.n3()
if o:
query['obj'] = _quoteLiteral(o.n3())
# o.n3()
# quote_plus(o.n3().encode("utf-8"))
if context:
# TODO FIXME what about bnodes like _:adf23123
query['context'] = "<%s>" % context
if query:
url = url + "?" + urlencode(query)
return url
def load(self, file_name, if_exists=ACTION_SKIP, method=METHOD_UPLOAD, batch_size=20):
file_path = os.path.abspath(file_name)
exists = os.path.isfile(file_path)
print("Trying to load {}".format(file_path))
if not exists:
return False, "{} file not found".format(file_path)
_ , file_extension = os.path.splitext(file_path)
file_extension = file_extension.strip('.')
if file_extension not in ['nt', 'ttl']:
return False, "Invalid extension {} : only ttl and nt files are supported now".format(file_extension)
# Load the TTL file
g = Graph()
g.parse(file_path, format=file_extension)
print("Parsed. Size {}".format(len(g)))
# Check if it contains the mandatory dataset definition
# TODO: This should be expanded to check also if author name,
# title, description, and licence are defined
results = g.triples( (None, RDF.type, DATASET_TYPE_URI) )
found = False
dataset_uri = None
for triple in results:
if not found:
found = True
dataset_uri, _ , _ = triple
else :
return False, "Multiple declarations of type {} . Only one expected".format(DATASET_TYPE_URI)
if not found:
return False, "Missing declarations of type {} . One expected".format(DATASET_TYPE_URI)
try:
if not self.exists(dataset_uri):
print("Creating graph {}".format(dataset_uri))
success, message = self.create(dataset_uri)
if not success:
return False, message
else :
print("Dataset {} exists".format(dataset_uri))
if if_exists == ACTION_SKIP:
return True, "Skipping import"
if if_exists == ACTION_DELETE:
success, message = self.delete(dataset_uri)
if not success:
return False, message
success, message = self.create(dataset_uri)
if not success:
return False, message
print("Deleted contents of {}".format(dataset_uri))
except Unauthorized as err :
return False, "Failed to connect to the data: access not allowed"
except urllib.error.HTTPError as err:
return False, "Failed to connect to the data: {}".format(err)
success, message = True, "OK"
if method == METHOD_UPLOAD and len(g) < 1000:
print("Uploading contents of {}".format(file_path))
files = { 'file': (os.path.basename(file_path), open(file_path,'rb'), 'text/turtle'), }
response = requests.post("{}?graph={}".format(self.upload_endpoint, dataset_uri),
files=files,
auth=(self.endpoint_user, self.endpoint_pwd)
)
success, message = response.status_code in [200, 201], response.text
else :
print("Serializing iterative insertions")
triples=[]
count_inserted=0
for sb,pr,obj in g:
if isinstance(obj, Literal):
triples.append("{} {} {}".format(sb.n3(), pr.n3(), _quoteLiteral(obj)))
else:
triples.append("{} {} {}".format(sb.n3(), pr.n3(), obj.n3()))
if len(triples) >= batch_size:
success, message = self.insert(dataset_uri, triples)
if not success:
return False, message
count_inserted+=1
triples=[]
if count_inserted%10 == 1:
print("Inserted {} ".format(count_inserted*batch_size))
sleep(3)
if len(triples) > 0:
success, message = self.insert(dataset_uri, triples)
return success, message
