def load_default_namespaces(directory):
print headings(
"LOADING DEFAULT NAMESPACES TO STARDOG [{}]".format(stardog_db))
if path.isdir(directory) is False:
return "\n>>> [{}] IS NOT A DIRECTORY ".format(directory)
f_path = path.join(directory,
"namespace.bat" if Ut.is_windows() else "namespace.sh")
# PLATFORM DEPENDENT CMD
if Ut.is_windows():
cmd = namespaces.format("call ", Svr.settings[St.stardog_uri])
else:
cmd = namespaces.format(
stardog_bin,
Svr.settings[St.stardog_uri],
)
# EXECUTE THE CMD
result = Ut.run_cdm(cmd, f_path, delete_after=True, output=False)
# DISPLAY THE FINAL RETURN
print "Finished with: {}".format(result)
def query_graph_metadata(graph):
print headings("QUERYING STARDOG FOR THE GENERIC METADATA OF A GRAPH")
print "{:12} : {}".format("INPUT GRAPH ", graph)
graph = main_alignment(graph)
print "{:12} : {}".format("MAIN GRAPH", graph)
qry = std_queries["metadata"].format(graph)
result = query(qry)
# print result
return result
def add_namespace(namespace, uri):
print headings("ADDING A NAMESPACE")
print "NAMESPACE LABEL:".format(namespace)
print "NAMESPACE URIS :".format(uri)
# PLATFORM DEPENDENT CMD
try:
cmd = "stardog namespace add {} --prefix {} --uri {}".format(
Svr.settings[St.stardog_uri], namespace, uri)
return subprocess.check_output(cmd, shell=True)
except ValueError:
traceback.print_exc()
def stardog_query_list():
"""""" """""" """""" """""" """""" """""" """""" """""" """""" """
# QUERYING STARDOG FOR THE CURRENT LIST OF QUERIES
""" """""" """""" """""" """""" """""" """""" """""" """""" """"""
print headings("QUERYING STARDOG FOR THE CURRENT LIST OF QUERIES")
try:
cmd = stardog_cmds["query_list"].format(stardog_bin, stardog_address)
remove = "{}".format(stardog_bin)
print "{:12} : {}".format("STARDOG COMMAND",
cmd.replace("\"", "").replace(remove, ""))
return subprocess.check_output(cmd, shell=True)
except Exception as err:
return err
def stardog_query_kill(query_id):
"""""" """""" """""" """""" """""" """""" """""" """""" """""" """
# TERMINATING A SPECIFIC QUERY BASED ON ID
""" """""" """""" """""" """""" """""" """""" """""" """""" """"""
print headings("TERMINATING A SPECIFIC QUERY BASED ON ID")
try:
cmd = stardog_cmds["query_kill"].format(stardog_bin, stardog_address,
query_id)
remove = "{}".format(stardog_bin)
print "{:12} : {}".format("STARDOG COMMAND",
cmd.replace("\"", "").replace(remove, ""))
return subprocess.check_output(cmd, shell=True)
except Exception as err:
return err
def query_generic(graph, limit=100):
"""""" """""" """""" """""" """""
# QUERRYING STARDOG
""" """""" """""" """""" """""" ""
print headings("GENERIC QUERY FOR STARDOG")
try:
cmd = std_queries["query_generic"].format(stardog_bin, stardog_uri,
graph, limit)
remove = "\"{}stardog\" query {} \"".format(stardog_bin, stardog_uri)
print "{:12} : {}".format("QUERY", cmd[0:-1].replace(remove, ""))
return subprocess.check_output(cmd, shell=True)
except Exception as err:
traceback.print_exc()
return err
def stardog_data_add_folder(folder_path,
named_graph=None,
database=None,
add=True,
fies_format="trig",
activated=False):
if activated is False:
message = "THE FUNCTION [stardog_data_add_folder] IS NOT ACTIVATED"
print message
return message
if fies_format.strip() == "ttl" and named_graph is None:
return "The named graph is required for loading your data."
"""""" """""" """""" """""" """""" """""" """""" """
# ADDING DATA TO STARDOG
""" """""" """""" """""" """""" """""" """""" """"""
print headings("ADDING DATA TO STARDOG FROM A FOLDER")
if database is None:
database = stardog_uri
if named_graph is not None:
graph = "-g {}".format(named_graph.strip())
else:
graph = ""
if folder_path.strip().endswith(path.sep) is False:
folder_path = "{}{}".format(folder_path, path.sep)
add_remove = "add" if add is True else "remove"
try:
cmd = stardog_cmds["data_add_folder"].format(stardog_bin, add_remove,
database, graph,
folder_path, fies_format)
cmd = cmd.replace("\\", "/")
remove = "{}".format(stardog_bin)
print "{:12} : {}".format("STARDOG COMMAND",
cmd.replace("\"", "").replace(remove, ""))
result = subprocess.check_output(cmd, shell=True)
print result
return result
except Exception as err:
traceback.print_exc()
return err
def query(query_search):
"""""" """""" """""" """""" """""
# QUERYING STARDOG
""" """""" """""" """""" """""" ""
print headings("QUERYING STARDOG")
try:
cmd = std_queries["query"].format(stardog_bin, stardog_uri,
query_search)
remove = "\"{}stardog\" query {} \"".format(stardog_bin, stardog_uri)
print "{:12} : {}".format("QUERY", cmd[0:-1].replace(remove, ""))
cmd = cmd.replace("\n", "")
return subprocess.check_output(cmd, shell=True)
except Exception as err:
traceback.print_exc()
return err
def stardog_query_status(query_id):
"""""" """""" """""" """""" """""" """""" """""" """""" """""" """""" """""" """
# ASSESSING THE STATUS OF A SPECIFIC CURRENTLY RUNNING QUERY
""" """""" """""" """""" """""" """""" """""" """""" """""" """""" """""" """"""
print headings(
"ASSESSING THE STATUS OF A SPECIFIC CURRENTLY RUNNING QUERY")
try:
cmd = stardog_cmds["query_status"].format(stardog_bin, stardog_address,
query_id)
remove = "{}".format(stardog_bin)
print "{:12} : {}".format("STARDOG COMMAND",
cmd.replace("\"", "").replace(remove, ""))
return subprocess.check_output(cmd, shell=True)
except Exception as err:
return err
def stardog_export_graph(file_path, graph, database=None):
"""""" """""" """""" """""" """""" """""" """""" """
# EXPORTING AN ENTIRE STARDOG GRAPH
""" """""" """""" """""" """""" """""" """""" """"""
print headings("EXPORTING AN ENTIRE STARDOG GRAPH")
if database is None:
database = stardog_uri
try:
cmd = stardog_cmds["export_graph"].format(stardog_bin, graph, database,
file_path)
remove = "{}".format(stardog_bin)
print "{:12} : {}".format("STARDOG COMMAND",
cmd.replace("\"", "").replace(remove, ""))
return subprocess.check_output(cmd, shell=True)
except Exception as err:
return err
def stardog_data_add_file(file_path,
graph=None,
database=None,
add=True,
activated=False):
if activated is False:
message = "THE FUNCTION [stardog_data_add_folder] IS NOT ACTIVATED"
print message
return message
"""""" """""" """""" """""" """""" """""" """""" """
# EXPORTING AN ENTIRE STARDOG DATABASE
""" """""" """""" """""" """""" """""" """""" """"""
print headings("ADDING DATA TO STARDOG FROM A FILE")
if database is None:
database = stardog_uri
if graph is not None:
graph = "-g {}".format(graph.strip())
add_remove = "add" if add is True else "remove"
try:
cmd = stardog_cmds["data_add_file"].format(stardog_bin, add_remove,
database, graph, file_path)
cmd = cmd.replace("\\", "/")
remove = "{}".format(stardog_bin)
print "{:12} : {}".format("STARDOG COMMAND",
cmd.replace("\"", "").replace(remove, ""))
result = subprocess.check_output(cmd, shell=True)
print result
return result
except Exception as err:
traceback.print_exc()
return err
def query_graph_search(search_exp):
print headings("QUERYING STARDOG FOR A SPECIFIC GRAPH")
qry = std_queries["graphs_search"].format(search_exp)
return query(qry)
def query_graph_properties(graph):
print headings("QUERYING STARDOG'S GRAPH PROPERTIES")
qry = std_queries["graph_properties"].format(graph)
return query(qry)
def query_graphs():
print headings("QUERYING STARDOG'S NAMED GRAPHS")
qry = std_queries["graph_all"]
return query(qry)
def __init__(self,
database,
is_trig,
file_to_convert,
separator,
entity_type,
rdftype=None,
subject_id=None,
embedded_uri=None,
field_metadata=None,
activated=False):
entity_type = entity_type.strip().replace(" ", "_")
database = database.strip().replace(" ", "_")
if activated is False:
print headings("The function [CSV init] has not been activated.")
return
specs = "\t{:18}: {}\n\t{:18}: {:18}\n\t{:18}: {}\n\t{:18}: [{}]\n\t{:18}: {}\n\t{:18}: {}".format(
"DATABASE NAME", database, "ENTITY-TYPE", entity_type,
"SUBJECT ID", subject_id, "COLUMN SEPARATOR", separator,
"TRIG FORMAT", str(is_trig), "FILE", file_to_convert)
print headings(
"CONVERTING TO RDF WITH THE FOLLOWING SPECS\n{}".format(specs))
# embedded_uri is an array of dictionaries.
# For each dictionary, We have:
# ID:Integer <- of the column that needs to became a uri
# reverse:Boolean <- to determine whether the URI needs to connect to the subject
# namespace:string e.g. http://risis.eu/ <- the namespace to assign to the new URI
# predicate:string It could be just the property (e.g. CityOf) or the URI (http://risis.eu/resource/CityOf)
print "RDF TYPE LIST : {}".format(rdftype)
print "SUBJECT ID : {}".format(subject_id)
"""
param database: name of the dataset
param is_trig: A boolean value indicating the format of the RDF file that will be generated.
param file_to_convert: Represents the oath of the CSV file that that is to be converted.
param separator: A character specifying the character used for value separation.
param subject_id: The index of the column identified to be used as the subject in the RDF file.
param entity_type: The name of the entity type starting by a capital character.
"""
self.embedded_uri = embedded_uri
self.fieldMetadata = field_metadata
if subject_id is None:
self.no_id(database,
is_trig,
file_to_convert,
separator,
entity_type,
rdftype,
activated=activated)
return
bom = ''
_file = ""
self.errorCount = 0
self.rdftype = rdftype
self.subjectID = subject_id # -> int The index of the attribute to use as identification
self.inputPath = file_to_convert # -> string The out file path
self.pvFormat = u"" # -> string Representing RDF triple format to use for formatting Predicate_Value
self.risisOntNeutral = "riClass:Neutral" # Prefix for Neutrality
self.lastColumn = 0 # -> int The last attribute index
self.longestHeader = 0 # -> int The number of characters in the longest attribute
# self.data_prefix = entity_type.lower().replace(' ', '_')
self.data_prefix = "resource" if entity_type is None else entity_type.strip(
).lower().replace(' ', '_')
'''Replace unwanted characters -> #;:.-(–)—[']`=’/”{“}^@*+!~\,%'''
self.pattern = '[?&#;:%!~+`=’*.(\-)–\\—@\[' ',\\]`{^}“/”]'
try:
# Open the file to convert
# _file = codecs.open(self.inputPath, 'rb', encoding="utf-8")
_file = open(self.inputPath, 'rb')
except Exception as exception:
print "\n", exception
exit(1)
""" About BYTE ORDER MARK (BOM) """
self.first_line = _file.readline().strip()
if self.first_line.startswith(to_bytes(codecs.BOM_UTF8)):
for i in range(len(to_bytes(codecs.BOM_UTF8))):
bom += self.first_line[i]
self.first_line = self.first_line.replace(bom, '')
print u"[" + os.path.basename(
self.inputPath) + u"]", u"contains BOM."
# get the first line
# self.first_line = self.first_line.strip(u'\r\n')
print "\n\tTHIS IS THE HEADER STRING : ", self.first_line
# Get the attribute headers
# -> Array about the list of attributes in the csv file
self.csvHeader = self.extractor(self.first_line, separator)
self.csvHeaderLabel = self.extractor(self.first_line, separator)
print "\tTHIS IS THE HEADER LIST : ", self.csvHeader
print "\tTHE HEADER LIST IS OF SIZE : ", len(self.csvHeader)
""" 2. Get the last column ID. This allows to stop the loop before the end
whenever the identification column happens to be the last column"""
self.lastColumn = len(self.csvHeader) - 1
# This is no longer the case because we now keep the column used as reference
# if self.subjectID == self.lastColumn:
# self.lastColumn -= 1
""" 3. Get the attribute headers and make them URI ready"""
for i in range(0, len(self.csvHeader)):
self.csvHeader[i] = self.csvHeader[i].replace(' ', '_')
self.csvHeader[i] = re.sub(self.pattern, u"",
self.csvHeader[i].replace('&', "_and_"))
'''For every attribute composed of more than 1 word and separated by space,
stat the first word with lower case followed by the underscore character'''
# print self.csvHeader
new_header = ""
header_split = self.csvHeader[i].split()
if header_split is not None and len(header_split) > 0:
new_header = header_split[0].lower()
for j in range(1, len(header_split)):
new_header += u"_" + header_split[j]
self.csvHeader[i] = new_header
# print header_split
'''Get the size (number of characters) of the longest attribute'''
if self.longestHeader < len(self.csvHeader[i]):
self.longestHeader = len(self.csvHeader[i])
""" 4. Set the RDF triple formatter """
sub_position = 6
# vocab: takes 6 slots
pre_position = sub_position + self.longestHeader
self.pvFormat = u"{0:>" + u"{0}".format(
str(sub_position)) + u"} {1:" + u"{0}".format(
str(pre_position)) + u"} {2}"
# GENERATING THE RDF SCHEMA
schema = self.get_schema(entity_type=entity_type,
field_metadata=self.fieldMetadata)
# print schema
RDF.__init__(self,
input_path=self.inputPath,
database=database,
entity_type=entity_type,
is_trig=is_trig,
namespace=self.get_namespace(database),
schema=schema)
n = 0
""" Opening the named-graph
"""
self.open_trig(dataset_prefix)
""" Writing the rdf instances of the dataset
"""
while True:
n += 1
line = to_unicode(_file.readline())
if not line:
# WRITE THE BAT FILE
print "\n\nFILE LOCATION: {}", self.dirName
self.bat_file = bat(self.dirName, self.database)
""" Closing the named-graph by closing the turtle writer.
CAN POSSIBLY THROUGH AND EXCEPTION BY RDFLIB AFTER CHECKING THE FILE
"""
if self.isClosed is not True:
self.close_writer()
print '\nNo more line... Process ended at line > ' + str(n)
print 'Done with converting [' + file_to_convert + '] to RDF!!!'
_file.close()
break
# if n <= 5:
# print line
# pass
# if n <= 72:
# """ Proceed with the conversion """
# self.write_triples(to_unicode(line), separator, embedded_uri, self.fieldMetadata)
buffered = ""
while True:
print >> sys.stderr, '\r', "\tCURRENT LINE: {}".format(n),
record = line
if not record:
break
if buffered != "":
div = CSV.extractor(record,
separator,
content_delimiter='"')
if len(div) != len(self.csvHeader):
record = u"{}{}".format(buffered, record)
div = CSV.extractor(record, separator, content_delimiter='"')
if len(div) < len(self.csvHeader):
buffered = u"{}".format(record)
print u">>> Buffered: {}".format(buffered)
elif len(div) == len(self.csvHeader):
# print "\nLINE: {}".format(record.rstrip())
# print "SIZE: {}".format(len(div))
buffered = ""
# for item in div:
# print "{}".format(item)
""" Proceed with the conversion """
self.write_triples(div, embedded_uri, self.fieldMetadata)
break
elif len(div) > len(self.csvHeader):
print "\nERROR!!!!"
# REPORT ERROR IN THE CHARACTER SEPARATED VALUE FORMAT
if len(div) != len(self.csvHeader):
self.errorCount += 1
print "{:5} Record encoding error. Header: {} columns while Record: {} columns".format(
self.errorCount, len(self.csvHeader), len(div))
print "\t\t{:8}".format(div)
# print line
# PRINTING ITEMS
for i in range(0, len(div)):
print b"\t\t{} - {}".format(
i + 1, to_bytes(div[i]))
break
line = to_unicode(_file.readline())
print "\n"
def union(specs, activated=False):
if activated is False:
# logger.warning("THE FUNCTION IS NOT ACTIVATED")
print ("THE FUNCTION IS NOT ACTIVATED")
return {St.message: "THE FUNCTION IS NOT ACTIVATED.", St.error_code: 1, St.result: None}
# print "\nEXECUTING UNION SPECS" \
# "\n======================================================" \
# "========================================================"
print headings("EXECUTING UNION SPECS...")
"""
THE generate_lens_name FUNCTION RETURNS THE NAME OF THE UNION AND A
QUERY THAT ALLOWS TO ASk WHETHER THE LENS TO BE CREATED EXIST BY CHECKING
WHETHER THERE EXISTS A LENS WITH THE SAME COMPOSITION IN TERMS GRAPHS USED FOR THE UNION
"""
# SET THE NAME OF THE UNION-LENS
print "1. DATASETS:", len(specs[St.datasets])
for ds in specs[St.datasets]:
print "\t- {}".format(ds)
info = Lu.generate_lens_name(specs[St.datasets])
specs[St.lens] = "{}{}".format(Ns.lens, info["name"])
print "\n2. LENS: ", info["name"]
# CHECK WHETHER THE LENS EXISTS
check = run_checks(specs, info["query"])
if check[St.result] != "GOOD TO GO":
if check[St.message].__contains__("ALREADY EXISTS"):
Urq.register_lens(specs, is_created=False)
return check
# print "AFTER CHECK"
# PREPARATION FOR THE CREATION OF THE LENS
specs[St.lens_target_triples] = ""
specs[St.expectedTriples] = 0
specs[St.insert_query] = ""
lens = specs[St.lens]
source = "{}{}".format(Ns.tmpgraph, "load00")
message_2 = Ec.ERROR_CODE_8.replace("#", specs[St.lens])
count = -1
insert_ans = False
try:
# GO THROUGH THE LINKSETS/LENSES IN THE LENS
# 1-SUM UP THE EXPECTED NUMBER OF TRIPLES
# 2-GENERATE THE TRIPLES REPRESENTATION OF GHE GRAPHS COMPOSING THIS LENS
# 3-GENERATE THE INSERT QUERY FOR MOVING BOTH LINKSET AND SINGLETON GRAPHS TO THE UNION GRAPH
total_size = 0
# LOAD ALL GRAPHS IN LOAD00
specs[St.insert_query] += "DROP SILENT GRAPH <{}{}> ;\n".format(Ns.tmpgraph, "load00")
# ITERATE THROUGH THE PROVIDED GRAPHS
for linkset in specs[St.datasets]:
# print "TARGET: ", linkset
count += 1
# GET THE TOTAL NUMBER OF CORRESPONDENCE TRIPLES INSERTED
curr_triples = Qry.get_triples(linkset)
# PROBABLY THE LINKSET HAS NO SUCH PROPERTY " void:triples ?triples ."
if curr_triples is None:
curr_triples = Qry.get_triples_count(linkset)
total_size += int(curr_triples)
print "{} Contains {} triples".format(linkset, curr_triples)
if curr_triples is not None:
specs[St.expectedTriples] += int(curr_triples)
else:
# THE IS A PROBLEM WITH THE GRAPH FOR SEVERAL POSSIBLE REASONS
return {St.message: message_2.replace("\n", "<br/>"), St.error_code: 1, St.result: None}
# GENERATE TRIPLES OUT OF THE TARGETS
specs[St.lens_target_triples] += "\n\t void:target <{}> ;".format(linkset)
# GET THE INSERT QUERY
# BOTH THE LINKSET AND THE SINGLETONS ARE MOVED TO A SINGLE GRAPH
partial_query = Qry.q_copy_graph(source, source, linkset)
if count == 0:
specs[St.insert_query] += partial_query
else:
specs[St.insert_query] += " ;\n{}".format(partial_query)
# INTERSECTION MANIPULATION OVER THE UNION (SOURCE)
insert_query = union_insert_q(lens, source, specs[St.lens_name])
# print "manipulation:", manipulation
specs[St.insert_query] += " ;\n{}".format(insert_query)
# GENERATE THE LENS UNION
if activated is True:
# print specs[St.insert_query]
insert_ans = Qry.boolean_endpoint_response(specs[St.insert_query])
specs[St.triples] = Qry.get_namedgraph_size(lens, isdistinct=False)
if specs[St.triples] == "0":
message = Ec.ERROR_CODE_9
print message
# return None
return {St.message: message.replace("\n", "<br/>"), St.error_code: 1, St.result: None}
# CHECK WHETHER THE RESULT CONTAINS DUPLICATES
contains_duplicated = Qry.contains_duplicates(lens)
print "Contains Opposite Direction Duplicated:", contains_duplicated
# IF IT DOES, REMOVE THE DUPLICATES
if contains_duplicated is True:
# logger.warning("THE LENS CONTAINS DUPLICATES.")
print "THE LENS CONTAINS DUPLICATES."
Qry.remove_duplicates(lens)
# logger.warning("THE DUPLICATES ARE NOW REMOVED.")
print "THE DUPLICATES ARE NOW REMOVED."
print "Number of triples loaded : {}".format(total_size)
specs[St.triples] = Qry.get_namedgraph_size(lens, isdistinct=False)
print "\t>>> INSERTED: {}\n\t>>> INSERTED TRIPLES: {}".format(insert_ans, specs[St.triples])
print "Inserted : {}".format(specs[St.triples])
print "Removed : {}".format(total_size - int(specs[St.triples]))
# LOAD THE METADATA
# NOT GOOD AS THE LENS ALSO HAS A SINGLETON GRAPH
# inserted_correspondences = int(Qry.get_union_triples(lens))
inserted_correspondences = int(specs[St.triples])
# print "inserted_correspondences:", inserted_correspondences
specs[St.removedDuplicates] = specs[St.expectedTriples] - inserted_correspondences
metadata = Gn.union_meta(specs)
# print "METADATA:", metadata
meta_ans = Qry.boolean_endpoint_response(metadata)
print "\t>>> IS THE METADATA GENERATED AND INSERTED? {}".format(meta_ans)
construct_response = Qry.get_constructed_graph(specs[St.lens])
if construct_response is not None:
print "\t>>> WRITING TO FILE"
construct_response = construct_response.replace('{', "<{}>\n{{".format(specs[St.lens]), 1)
write_to_file(
graph_name=specs[St.lens_name], metadata=None, correspondences=construct_response, directory=DIRECTORY)
print "\tLens created as : ", specs[St.lens]
# REGISTER THE LINKSET
Urq.register_lens(specs, is_created=True)
# return specs[St.lens]
message = "THE LENS WAS CREATED as {}. " \
"With initially {} triples loaded, {} duplicated triples were found and removed.".\
format(specs[St.lens], total_size, total_size - int(specs[St.triples]))
print "\t*** JOB DONE! ***"
return {St.message: message, St.error_code: 0, St.result: specs[St.lens]}
except Exception as err:
# logger.warning(err)
if insert_ans == "true":
"DROP THE INSERTED UNION"
drop_linkset(lens, activated=True)
print "ERROR IN UNION LENS CREATION:", err
return {St.message: ERROR_CODE_11, St.error_code: 11, St.result: None}
