def main_alignment(alignment):
# ****************************************************************************
# GIVEN AN ALIGNMENT, RETURN THE MAIN ONE
# ****************************************************************************
try:
# LOCAL NAME OF THE GRAPH
name = Ut.get_uri_local_name_plus(alignment)
print "{:12} : {}".format("LOCAL NAME", name)
query_search = std_queries["graphs_search"].format(name)
response = Qry.sparql_xml_to_matrix(query_search)
results = response["result"]
if results is not None:
for i in range(1, len(results)):
if results[i][0].__contains__("singletons") is False:
return results[i][0]
if str(alignment).__contains__(Ns.singletons):
return str(alignment).replace(Ns.singletons, Ns.linkset)
else:
return alignment
except ValueError:
traceback.print_exc()
return alignment
def main_alignment(alignment):
# LOCAL NAME OF THE GRAPH
name = Ut.get_uri_local_name_plus(alignment)
print "{:12} : {}".format("LOCAL NAME", name)
query = std_queries["graphs_search"].format(name)
response = Qry.sparql_xml_to_matrix(query)
results = response["result"]
if results is not None:
for i in range(1, len(results)):
if results[i][0].__contains__("singletons") is False:
return results[i][0]
if str(alignment).__contains__(Ns.singletons):
return str(alignment).replace(Ns.singletons, Ns.linkset)
else:
return alignment
def export_alignment_all(alignment, directory=None, limit=5000):
directory = os.path.join(directory, "")
print directory
if os.path.isdir(os.path.dirname(directory)) is False or os.path.exists(
directory) is False:
print "CREATING THE DIRECTORY"
os.mkdir(os.path.dirname(directory))
# COMMENT THE LINKSET OIT IF IT IS EQUAL TO NONE
# This function returns all the links + some metadata about the alignment.
# METADATA: source dataset, target dataset and mechanism
use = alignment
alignment = str(alignment).strip()
row_alignment = alignment
alignment = alignment if Ut.is_nt_format(
alignment) is True else "<{}>".format(alignment)
# ****************************************************
# 1. GET THE METADATA OF THE ALIGNMENT: THE QUERY
# ****************************************************
meta = """
PREFIX ll: <{0}>
CONSTRUCT {{ {1} ?y ?z. ?z ?p ?o . }}
WHERE
{{
{1} ?y ?z .
OPTIONAL{{ ?z ?p ?o . }}
OPTIONAL{{ ?O ?Q ?R . }}
}} order by ?y
""".format(Ns.alivocab, alignment)
# print meta
# GET THE METADATA OF THE ALIGNMENT: RUN THE QUERY
meta_construct = Qry.endpointconstruct(meta, clean=False)
meta_construct = meta_construct.replace("{", "").replace("}", "")
with open(os.path.join(directory, "metadata.ttl"), "wb") as metadata:
metadata.write(meta_construct)
# print meta_construct
# ****************************************************
# 2. GET THE CORRESPONDENCES OF THE LINKSET
# ****************************************************
# CONSTRUCT QUERY FOR EXTRACTING HE CORRESPONDENCES
comment = "" if limit else "#"
query = """
PREFIX ll: <{}>
CONSTRUCT {{ ?x ?y ?z }}
WHERE
{{
GRAPH {}
{{
?x ?y ?z
}}
}} order by ?x {}LIMIT {}
""".format(Ns.alivocab, alignment, comment, limit)
# print query
# FIRE THE CONSTRUCT FOR CORRESPONDENCES AGAINST THE TRIPLE STORE
alignment_construct = Qry.endpointconstruct(query, clean=False)
if alignment_construct:
alignment_construct = alignment_construct.replace(
"{", "{}\n{{".format(alignment))
# print alignment_construct
with open(os.path.join(directory, "linkset.trig"), "wb") as links:
links.write(alignment_construct)
# ****************************************************
# 3. GET THE METADATA CORRESPONDENCES' PREDICATES
# ****************************************************
singleton_graph_uri = Ut.from_alignment2singleton(alignment)
singleton_query = """
PREFIX ll: <{0}>
PREFIX singletons: <{1}>
CONSTRUCT {{ ?predicate ?x ?y }}
WHERE
{{
{{
SELECT ?predicate
{{
GRAPH {2}
{{
?subject ?predicate ?object
}}
}} order by ?x {3}LIMIT {4}
}}
GRAPH {5}
{{
?predicate ?x ?y
}}
}}
""".format(Ns.alivocab, Ns.singletons, alignment, comment, limit,
singleton_graph_uri)
# print singleton_query
# FIRE THE CONSTRUCT FOR SINGLETON AGAINST THE TRIPLE STORE
singleton_construct = Qry.endpointconstruct(singleton_query, clean=False)
if singleton_construct:
singleton_construct = singleton_construct.replace(
"{", "{}\n{{".format(singleton_graph_uri))
# print singleton_construct
with open(os.path.join(directory, "singletons.trig"), "wb") as singletons:
singletons.write(singleton_construct)
# LOAD THE METADATA USING RDFLIB
sg = rdflib.Graph()
sg.parse(data=meta_construct, format="turtle")
# EXTRACT FROM THE RESPONSE: THE SOURCE AND TARGET DATASETS AND THE ALIGNMENT
sbj = rdflib.URIRef(use)
triples_uri = rdflib.URIRef("http://rdfs.org/ns/void#triples")
# EXTRACT THE ALIGNMENT TYPE
triples = ""
for item in sg.objects(sbj, triples_uri):
triples = item
print "TRIPLES: ", triples
if alignment_construct is not None:
links = "### TRIPLE COUNT: {}\n### LINKSET: {}\n".format(
triples, alignment) + alignment_construct
links = links.replace("{", "").replace("}", "")
message = "You have just downloaded the graph [{}] which contains [{}] correspondences. ".format(
row_alignment, triples)
host = Svr.settings[St.stardog_host_name]
endpoint = b"http://{}/annex/{}/sparql/query?".format(
host, Svr.settings[St.database])
local_name = Ut.get_uri_local_name_plus(alignment)
file_at_parent_directory = os.path.join(
os.path.abspath(os.path.join(directory, os.pardir)),
"{}.zip".format(local_name))
zipped_file = Ut.zip_folder(directory,
output_file_path=file_at_parent_directory)
print "\t>>> THE ZIPPED FILE IS LOCATED AT:\n\t\t- {}".format(zipped_file)
# result = result
# print result
print "Done with graph: {}".format(alignment)
# return {'result': {
# "generic_metadata": meta_construct,
# 'specific_metadata': singleton_construct,
# 'data': alignment_construct}, 'message': message}
return {'result': zipped_file, 'message': message}
def intersection_extended(specs, lens_name, display=False):
# print Ut.headings("EXTENDED INTERSECTION")
inter = ""
insert = Buffer.StringIO()
insert_sing = Buffer.StringIO()
model_1 = """
### ABOUT {0}
GRAPH <{0}>
{{
{1}
}}
"""
model_2 = """
### {2}. ABOUT {0}
GRAPH <{0}>
{{
?{1} ?pred_{2} ?{3} .
}}
### SINGLETONS
GRAPH <{4}>
{{
?pred_{2} ?x_{2} ?y_{2} .
}}"""
count_graph = 1
for graph in specs[St.datasets]:
query = """
PREFIX void: <{}>
PREFIX bdb: <{}>
SELECT distinct ?subTarget ?objTarget ?subjectEntityType ?objectEntityType
{{
<{}>
#void:target*/(void:subjectsTarget|void:objectsTarget)* ?x ;
void:target*/(void:subjectsTarget|void:objectsTarget)* ?x .
?x
void:subjectsTarget ?subTarget ;
void:objectsTarget ?objTarget ;
bdb:subjectsDatatype ?subjectEntityType ;
bdb:objectsDatatype ?objectEntityType .
FILTER NOT EXISTS {{ ?subTarget a void:Linkset }}
FILTER NOT EXISTS {{ ?objTarget a void:Linkset }}
}}""".format(Ns.void, Ns.bdb, graph)
# print "INTERSECTION QUERY:", query
response = sparql_xml_to_matrix(query)
if display:
print "INTERSECTION QUERY:", query
# print "\nGRAPH:", graph
# print "RESPONSE:", response
# exit(0)
if response:
targets = response[St.result]
# IF THE RESULT HAS MORE THAN
# print "LENGTH:", len(targets)
if targets is not None and len(targets) > 2:
union = ""
for i in range(1, len(targets)):
append = "UNION" if i < len(targets) - 1 else ""
tab = "" if i == 1 else ""
src = Ut.get_uri_local_name(targets[i][0])
trg = Ut.get_uri_local_name(targets[i][1])
if src[0].isdigit():
src = "D{}".format(src)
if trg[0].isdigit():
trg = "D{}".format(trg)
src_TYPE = Ut.get_uri_local_name(targets[i][2])
trg_TYPE = Ut.get_uri_local_name(targets[i][3])
src_variable = "{}_{}_1".format(src, src_TYPE[short:])
if src == trg and src_TYPE == trg_TYPE:
trg_variable = "{}_{}_2".format(trg, trg_TYPE[short:])
else:
trg_variable = "{}_{}_1".format(trg, trg_TYPE[short:])
union += "\n\t\t{0}{{ ?{1} ?predicate_{2} ?{3} . }} {4}".format(
tab, src_variable, i, trg_variable, append)
union = model_1.format(graph, union)
# print "UNION:", union
inter += union
# ONLY TWO TARGETS
elif targets and len(targets) == 2:
src = Ut.get_uri_local_name(targets[1][0])
trg = Ut.get_uri_local_name(targets[1][1])
if src[0].isdigit():
src = "D{}".format(src)
if trg[0].isdigit():
trg = "D{}".format(trg)
src_TYPE = Ut.get_uri_local_name(targets[1][2])
trg_TYPE = Ut.get_uri_local_name(targets[1][3])
src_variable = "{}_{}_1".format(src, src_TYPE[short:])
if src == trg and src_TYPE == trg_TYPE:
trg_variable = "{}_{}_2".format(trg, trg_TYPE[short:])
else:
trg_variable = "{}_{}_1".format(trg, trg_TYPE[short:])
inter += model_2.format(
graph, src_variable, count_graph, trg_variable,
"{}{}".format(Ns.singletons,
Ut.get_uri_local_name_plus(graph)))
insert.write("\t\t?{} ?pred_{} ?{} .\n".format(
src_variable, count_graph, trg_variable))
insert_sing.write(
"\t\t?pred_{0} ?x_{0} ?y_{0}.\n".format(count_graph))
count_graph += 1
# print inter
# exit(0)
insert_query = """INSERT\n{{
### LINKS
GRAPH <{5}{4}>
{{\n{1}\t}}
### METADATA
GRAPH <{6}{4}>
{{\n{3}\t}}\n}}\nWHERE\n{{{2}\n}}
""".format("", insert.getvalue(), inter, insert_sing.getvalue(), lens_name,
Ns.lens, Ns.singletons)
return insert_query
