def del_vocab_from_creator(userid, vocab):
if not os.path.isfile(os.path.join(ag.creatorsdir, '%s.rdf'%userid)):
return False
graph = Graph()
graph.parse(os.path.join(ag.creatorsdir, '%s.rdf'%userid))
vocab_uri = URIRef("http://vocab.ox.ac.uk/%s"%vocabprefix)
for s, p, o in graph.triples((URIRef(vocab_uri), namespaces['dcterms']['mediator'], None)):
graph.remove((s, p, o))
rdf_str = None
rdf_str = graph.serialize()
f = codecs.open(creatorfile, 'w', 'utf-8')
f.write(rdf_str)
f.close()
return True
def del_vocab_from_creator(userid, vocab):
if not os.path.isfile(os.path.join(ag.creatorsdir, '%s.rdf' % userid)):
return False
graph = Graph()
graph.parse(os.path.join(ag.creatorsdir, '%s.rdf' % userid))
vocab_uri = URIRef("http://vocab.ox.ac.uk/%s" % vocabprefix)
for s, p, o in graph.triples(
(URIRef(vocab_uri), namespaces['dcterms']['mediator'], None)):
graph.remove((s, p, o))
rdf_str = None
rdf_str = graph.serialize()
f = codecs.open(creatorfile, 'w', 'utf-8')
f.write(rdf_str)
f.close()
return True
class TestSPARQLToldBNodes(unittest.TestCase):
def setUp(self):
NS = u"http://example.org/"
self.graph = ConjunctiveGraph()
self.graph.parse(StringInputSource("""
@prefix : <http://example.org/> .
@prefix rdf: <%s> .
@prefix rdfs: <%s> .
[ :prop :val ].
[ a rdfs:Class ]."""%(RDF.RDFNS,RDFS.RDFSNS)), format="n3")
def testToldBNode(self):
for s,p,o in self.graph.triples((None,RDF.type,None)):
pass
query = """SELECT ?obj WHERE { %s ?prop ?obj }"""%s.n3()
print query
rt = self.graph.query(query)
self.failUnless(len(rt) == 1,"BGP should only match the 'told' BNode by name (result set size: %s)"%len(rt))
class sparql_funcs():
def __init__(self):
self.g = Graph('IOMemory')
#self.endpoint = "http://www.opencorrespondence.org/data/endpoint/rdf"
#self.g.bind('geo', geo)
def find_places(self):
'''
Function to get the distinct locations mentioned in the headers of the letters.
These are the locations from which Dickens wrote.
TODO: Parsing the letters to get the places mentioned in them
'''
row = set()
o = OFS()
for b in o.list_buckets():
endpoint = o.get_stream(b, "endpoint")
self.g.parse(endpoint)
for s, _, n in self.g.triples((None, dublin_core['title'], None)):
loc_key = urllib.unquote(
n.replace("http://www.opencorrespondence.org/place/resource/",
"").replace("/rdf", ""))
row.add(self.tidy_location(loc_key))
return row
def tidy_location(self, location):
'''
Function to tidy up some of the places where they refer to the same place
TODO: prob need some language processing to make this scalable
'''
ret_location = ''
if location == 'Office Of "household Words,':
ret_location = "Household Words"
elif location == '"household Words" Office':
ret_location = "Household Words"
elif location == '"household Words"':
ret_location = "Household Words"
elif location == 'H. W. Office':
ret_location = "Household Words"
elif location == '"household Words,':
ret_location = "Household Words"
elif location == '"all The Year Round" Office':
ret_location = "All The Year Round"
elif location == 'Office Of "all The Year Round,':
ret_location = "All The Year Round"
elif location == "Gad's Hill Place":
ret_location = "Gads Hill"
elif location == "Gad's Hill":
ret_location = "Gads Hill"
elif location == "Gad's Hill Place, Higham":
ret_location = "Gads Hill"
elif location == "Tavistock House, Tavistock Square":
ret_location = "Tavistock House"
elif location == "London, Tavistock House":
ret_location = "Tavistock House"
elif location == "Tavistock House, London":
ret_location = "Tavistock House"
else:
if "U.s." in location:
location = str(location).replace("U.s", "")
ret_location = str(location).replace(".", "")
return ret_location
def find_correspondents(self):
'''
Function to get the distinct locations mentioned in the headers of the letters.
These are the locations from which Dickens wrote.
TODO: Parsing the letters to get the places mentioned in them
'''
row = set()
self.g.parse(self.endpoint)
for s, _, n in self.g.triples((None, letter['correspondent'], None)):
loc_key = urllib.unquote(
n.replace(
"http://www.opencorrespondence.org/correspondent/resource/",
"").replace("/rdf", ""))
row.add(loc_key)
return row
def get_abstract(self, resource_id):
self.g.parse('http://dbpedia.org/resource/'.resource_id)
q = '''
SELECT *
WHERE
{
?x dbpedia:abstract ?abstract .
FILTER (lang(?abstract) = 'en')
}
'''
for row in self.g.query(
q,
initNs=dict(dbpedia=Namespace("http://dbpedia.org/ontology/")),
initBindings={}):
return row[1]
def query_dates(self, author):
'''query to identify individual dates to a correspondent'''
q = '''
SELECT ?date
FROM <http://localhost:5000/data/endpoint/rdf>
WHERE {
?r dc:subject \'''' + author + '''\' .
?r dc:date ?date.
}
'''
dates = []
for row in self.g.query(
q,
initNs=dict(letter=Namespace(
"http://www.opencorrespondence.org/schema#"),
dc=Namespace("http://purl.org/dc/elements/1.1/")),
initBindings={}):
date = str(row[0]).split('-')
if date[0][1:].isdigit():
dates.append(date[0])
print dates
dic = {}
for dt in dates:
dic[dt] = dates.count(dt)
return dic
def testN3Store(store="default", configString=None):
g = ConjunctiveGraph(store=store)
if configString:
g.destroy(configString)
g.open(configString)
g.parse(StringInputSource(testN3), format="n3")
print g.store
try:
for s, p, o in g.triples((None, implies, None)):
formulaA = s
formulaB = o
assert type(formulaA) == QuotedGraph and type(formulaB) == QuotedGraph
a = URIRef('http://test/a')
b = URIRef('http://test/b')
c = URIRef('http://test/c')
d = URIRef('http://test/d')
v = Variable('y')
universe = ConjunctiveGraph(g.store)
#test formula as terms
assert len(list(universe.triples((formulaA, implies, formulaB)))) == 1
#test variable as term and variable roundtrip
assert len(list(formulaB.triples((None, None, v)))) == 1
for s, p, o in formulaB.triples((None, d, None)):
if o != c:
assert isinstance(o, Variable)
assert o == v
s = list(universe.subjects(RDF.type, RDFS.Class))[0]
assert isinstance(s, BNode)
assert len(list(universe.triples((None, implies, None)))) == 1
assert len(list(universe.triples((None, RDF.type, None)))) == 1
assert len(list(formulaA.triples((None, RDF.type, None)))) == 1
assert len(list(formulaA.triples((None, None, None)))) == 2
assert len(list(formulaB.triples((None, None, None)))) == 2
assert len(list(universe.triples((None, None, None)))) == 3
assert len(
list(formulaB.triples((None, URIRef('http://test/d'), None)))) == 2
assert len(
list(universe.triples((None, URIRef('http://test/d'), None)))) == 1
#context tests
#test contexts with triple argument
assert len(list(universe.contexts((a, d, c)))) == 1
#Remove test cases
universe.remove((None, implies, None))
assert len(list(universe.triples((None, implies, None)))) == 0
assert len(list(formulaA.triples((None, None, None)))) == 2
assert len(list(formulaB.triples((None, None, None)))) == 2
formulaA.remove((None, b, None))
assert len(list(formulaA.triples((None, None, None)))) == 1
formulaA.remove((None, RDF.type, None))
assert len(list(formulaA.triples((None, None, None)))) == 0
universe.remove((None, RDF.type, RDFS.Class))
#remove_context tests
universe.remove_context(formulaB)
assert len(list(universe.triples((None, RDF.type, None)))) == 0
assert len(universe) == 1
assert len(formulaB) == 0
universe.remove((None, None, None))
assert len(universe) == 0
g.store.destroy(configString)
except:
g.store.destroy(configString)
raise
def testN3Store(store="default", configString=None):
g = ConjunctiveGraph(store=store)
if configString:
g.destroy(configString)
g.open(configString)
g.parse(StringInputSource(testN3), format="n3")
print g.store
try:
for s,p,o in g.triples((None,implies,None)):
formulaA = s
formulaB = o
assert type(formulaA)==QuotedGraph and type(formulaB)==QuotedGraph
a = URIRef('http://test/a')
b = URIRef('http://test/b')
c = URIRef('http://test/c')
d = URIRef('http://test/d')
v = Variable('y')
universe = ConjunctiveGraph(g.store)
#test formula as terms
assert len(list(universe.triples((formulaA,implies,formulaB))))==1
#test variable as term and variable roundtrip
assert len(list(formulaB.triples((None,None,v))))==1
for s,p,o in formulaB.triples((None,d,None)):
if o != c:
assert isinstance(o,Variable)
assert o == v
s = list(universe.subjects(RDF.type, RDFS.Class))[0]
assert isinstance(s,BNode)
assert len(list(universe.triples((None,implies,None)))) == 1
assert len(list(universe.triples((None,RDF.type,None)))) ==1
assert len(list(formulaA.triples((None,RDF.type,None))))==1
assert len(list(formulaA.triples((None,None,None))))==2
assert len(list(formulaB.triples((None,None,None))))==2
assert len(list(universe.triples((None,None,None))))==3
assert len(list(formulaB.triples((None,URIRef('http://test/d'),None))))==2
assert len(list(universe.triples((None,URIRef('http://test/d'),None))))==1
#context tests
#test contexts with triple argument
assert len(list(universe.contexts((a,d,c))))==1
#Remove test cases
universe.remove((None,implies,None))
assert len(list(universe.triples((None,implies,None))))==0
assert len(list(formulaA.triples((None,None,None))))==2
assert len(list(formulaB.triples((None,None,None))))==2
formulaA.remove((None,b,None))
assert len(list(formulaA.triples((None,None,None))))==1
formulaA.remove((None,RDF.type,None))
assert len(list(formulaA.triples((None,None,None))))==0
universe.remove((None,RDF.type,RDFS.Class))
#remove_context tests
universe.remove_context(formulaB)
assert len(list(universe.triples((None,RDF.type,None))))==0
assert len(universe)==1
assert len(formulaB)==0
universe.remove((None,None,None))
assert len(universe)==0
g.store.destroy(configString)
except:
g.store.destroy(configString)
raise
from __future__ import division
import sys, time
sys.path.append('pytc-0.7/build/lib.linux-x86_64-2.5')
sys.path.insert(0, 'rdflib/build/lib.linux-x86_64-2.5')
from rdflib.Graph import ConjunctiveGraph
from rdflib import Namespace, Literal
EX = Namespace("http://example.com/")
graph = ConjunctiveGraph('TokyoCabinet')
graph.open("tcgraph", create=True)
graph.add((EX['hello'], EX['world'], Literal("lit")))
print list(graph.triples((None, None, None)))
graph.close()
class sparql_funcs():
def __init__(self):
self.g = Graph('IOMemory')
#self.endpoint = "http://www.opencorrespondence.org/data/endpoint/rdf"
#self.g.bind('geo', geo)
def find_places(self):
'''
Function to get the distinct locations mentioned in the headers of the letters.
These are the locations from which Dickens wrote.
TODO: Parsing the letters to get the places mentioned in them
'''
row = set()
o = OFS()
for b in o.list_buckets():
endpoint = o.get_stream(b, "endpoint")
self.g.parse(endpoint)
for s,_,n in self.g.triples((None, dublin_core['title'], None)):
loc_key = urllib.unquote(n.replace("http://www.opencorrespondence.org/place/resource/", "").replace("/rdf",""))
row.add(self.tidy_location(loc_key))
return row
def tidy_location (self, location):
'''
Function to tidy up some of the places where they refer to the same place
TODO: prob need some language processing to make this scalable
'''
ret_location = '';
if location == 'Office Of "household Words,':
ret_location = "Household Words"
elif location== '"household Words" Office':
ret_location = "Household Words"
elif location== '"household Words"':
ret_location = "Household Words"
elif location== 'H. W. Office':
ret_location = "Household Words"
elif location == '"household Words,':
ret_location = "Household Words"
elif location == '"all The Year Round" Office':
ret_location = "All The Year Round"
elif location == 'Office Of "all The Year Round,':
ret_location = "All The Year Round"
elif location == "Gad's Hill Place":
ret_location = "Gads Hill"
elif location == "Gad's Hill":
ret_location = "Gads Hill"
elif location == "Gad's Hill Place, Higham":
ret_location = "Gads Hill"
elif location == "Tavistock House, Tavistock Square":
ret_location = "Tavistock House"
elif location == "London, Tavistock House":
ret_location = "Tavistock House"
elif location == "Tavistock House, London":
ret_location = "Tavistock House"
else:
if "U.s." in location:
location = str(location).replace("U.s", "")
ret_location = str(location).replace(".", "")
return ret_location
def find_correspondents(self):
'''
Function to get the distinct locations mentioned in the headers of the letters.
These are the locations from which Dickens wrote.
TODO: Parsing the letters to get the places mentioned in them
'''
row = set()
self.g.parse(self.endpoint)
for s,_,n in self.g.triples((None, letter['correspondent'], None)):
loc_key = urllib.unquote(n.replace("http://www.opencorrespondence.org/correspondent/resource/", "").replace("/rdf", ""))
row.add(loc_key)
return row
def get_abstract (self, resource_id):
self.g.parse('http://dbpedia.org/resource/'.resource_id)
q = '''
SELECT *
WHERE
{
?x dbpedia:abstract ?abstract .
FILTER (lang(?abstract) = 'en')
}
'''
for row in self.g.query(q,
initNs=dict(dbpedia=Namespace("http://dbpedia.org/ontology/")),
initBindings={}):
return row[1]
def query_dates(self, author):
'''query to identify individual dates to a correspondent'''
q = '''
SELECT ?date
FROM <http://localhost:5000/data/endpoint/rdf>
WHERE {
?r dc:subject \'''' + author + '''\' .
?r dc:date ?date.
}
'''
dates = []
for row in self.g.query(q,
initNs=dict(letter=Namespace("http://www.opencorrespondence.org/schema#"), dc=Namespace("http://purl.org/dc/elements/1.1/")),
initBindings={}):
date = str(row[0]).split('-')
if date[0][1:].isdigit():
dates.append(date[0])
print dates
dic = {}
for dt in dates:
dic[dt] = dates.count(dt)
return dic
