def delete_datasets(self):
"""delete the datasets from the database and the triplestore"""
sparql = SparqlQueryLauncher(self.app, self.session)
for dataset in self.datasets:
# Delete from triplestore
if dataset.graph_name:
Utils.redo_if_failure(self.log, 3, 1, sparql.drop_dataset,
dataset.graph_name)
# Delete from db
dataset.delete_from_db()
def delete_user_rdf(self, username):
"""Delete a user rdf graphs
Delete in DB, TS and filesystem
Parameters
----------
username : string
Username to delete
"""
tse = TriplestoreExplorer(self.app, self.session)
query_launcher = SparqlQueryLauncher(self.app, self.session)
graphs = tse.get_graph_of_user(username)
for graph in graphs:
Utils.redo_if_failure(self.log, 3, 1, query_launcher.drop_dataset, graph)
def delete_datasets(self):
"""delete the datasets from the database and the triplestore"""
sparql = SparqlQueryLauncher(self.app, self.session)
tse = TriplestoreExplorer(self.app, self.session)
for dataset in self.datasets:
# Delete from triplestore
if dataset.graph_name:
Utils.redo_if_failure(self.log, 3, 1, sparql.drop_dataset,
dataset.graph_name)
# Delete from db
dataset.delete_from_db()
# Uncache abstraction
tse.uncache_abstraction(public=dataset.public)
def rdfize(self, string, custom_namespace=None):
"""Rdfize a string
Return the literal if string is an url, else,
prefix it with askomics prefix
Parameters
----------
string : string
Term to rdfize
Returns
-------
rdflib.???
Rdfized term
"""
if Utils.is_valid_url(string):
return rdflib.URIRef(string)
elif ":" in string and len(string.split(":")) == 2:
prefix, val = string.split(":")
if prefix:
prefix_manager = PrefixManager(self.app, self.session)
namespace = prefix_manager.get_namespace(prefix)
if namespace:
return rdflib.URIRef("{}{}".format(namespace, val))
else:
# If not prefix, default to entity prefix
string = val
if custom_namespace:
return custom_namespace[self.format_uri(string)]
else:
return self.namespace_data[self.format_uri(string)]
def get_file_name(self):
"""Get a random file name
Returns
-------
string
file name
"""
name = Utils.get_random_string(10)
file_path = "{}/{}".format(self.upload_path, name)
# Make sure it is not in use already
while os.path.isfile(file_path):
name = Utils.get_random_string(10)
file_path = "{}/{}".format(self.upload_path, name)
return name
def create_reset_token(self, login):
"""Insert a token into the db
Parameters
----------
login : str
username or email
Returns
-------
str
The reset token
"""
token = "{}:{}".format(int(time.time()), Utils.get_random_string(20))
database_field = 'username'
if validate_email(login):
database_field = 'email'
database = Database(self.app, self.session)
query = """
UPDATE users
SET reset_token=?
WHERE {}=?
""".format(database_field)
database.execute_sql_query(query, (token, login))
return token
def replace_variables_in_blocks(self, var_to_replace):
"""Replace variables in blocks
Parameters
----------
var_to_replace : list of tuples
var to replace in block
"""
for var_source, var_target in var_to_replace:
# Interate throught blocks
for nblock, block in enumerate(self.triples_blocks):
# Iterate over sub-blocks
for nsblock, sblock in enumerate(block["sblocks"]):
# Iterate over triples
for ntriple, triple_dict in enumerate(sblock["triples"]):
for key, value in triple_dict.items():
if key != "optional":
self.triples_blocks[nblock]["sblocks"][nsblock]["triples"][ntriple][key] = value.replace(var_source, var_target)
for i, filtr in enumerate(sblock["filters"]):
self.triples_blocks[nblock]["sblocks"][nsblock]["filters"][i] = filtr.replace(var_source, var_target)
for i, value in enumerate(sblock["values"]):
self.triples_blocks[nblock]["sblocks"][nsblock]["values"][i] = value.replace(var_source, var_target)
self.triples_blocks[nblock]["sblocks"][nsblock]["triples"] = Utils.unique(self.triples_blocks[nblock]["sblocks"][nsblock]["triples"])
def update_apikey(self, user):
"""Create a new api key and store in the database
Parameters
----------
user : dict
The current user
Returns
-------
dict
error, error message and updated user
"""
error = False
error_message = ''
database = Database(self.app, self.session)
# get a new api key
new_apikey = Utils.get_random_string(20)
query = '''
UPDATE users SET
apikey=?
WHERE username=?
'''
database.execute_sql_query(query, (new_apikey, user['username']))
user['apikey'] = new_apikey
return {'error': error, 'error_message': error_message, 'user': user}
def get_uri_label(self, string):
"""Labelize a string
Try to extract a label from an URI
Parameters
----------
uri : string
Term to extract label from
Returns
-------
String
Label
"""
if Utils.is_valid_url(string):
string = string.rstrip("/")
if "/" in string:
end_term = string.split("/")[-1].rstrip("#")
if "#" in end_term:
end_term = end_term.split("#")[-1]
else:
end_term = string
elif ":" in string and len(string.split(":")) == 2:
end_term = string.split(":")[-1]
else:
end_term = string
return end_term
def integrate(self, public=False):
"""Integrate the file into the triplestore
Parameters
----------
public : bool, optional
Integrate in private or public graph
"""
sparql = SparqlQueryLauncher(self.app, self.session)
tse = TriplestoreExplorer(self.app, self.session)
self.public = public
method = self.settings.get('triplestore', 'upload_method')
# Load file into a RDF graph
self.graph_chunk.parse(self.path, format=self.type_dict[self.type])
# get metadata
self.set_metadata()
# Remove metadata from data
self.delete_metadata_location()
# insert metadata
sparql.insert_data(self.graph_metadata, self.file_graph, metadata=True)
if method == "load":
# write rdf into a tmpfile and load it
temp_file_name = 'tmp_{}_{}.{}'.format(Utils.get_random_string(5),
self.name,
self.rdf_extention)
# Try to load data. if failure, wait 5 sec and retry 5 time
Utils.redo_if_failure(self.log, 5, 5, self.load_graph,
self.graph_chunk, temp_file_name)
else:
# Insert
# Try to insert data. if failure, wait 5 sec and retry 5 time
Utils.redo_if_failure(self.log, 5, 5, sparql.insert_data,
self.graph_chunk, self.file_graph)
# Remove chached abstraction
tse.uncache_abstraction(public=self.public)
self.set_triples_number()
def set_graphs_and_endpoints(self,
entities=None,
graphs=None,
endpoints=None):
"""Get all public and private graphs containing the given entities
Parameters
----------
entities : list, optional
list of entity uri
"""
substrlst = []
filter_entity_string = ''
if entities:
for entity in entities:
substrlst.append("?entity_uri = <{}>".format(entity))
filter_entity_string = 'FILTER (' + ' || '.join(substrlst) + ')'
filter_public_string = 'FILTER (?public = <true>)'
if 'user' in self.session:
filter_public_string = 'FILTER (?public = <true> || ?creator = <{}>)'.format(
self.session["user"]["username"])
query = '''
SELECT DISTINCT ?graph ?endpoint
WHERE {{
?graph :public ?public .
?graph dc:creator ?creator .
GRAPH ?graph {{
?graph prov:atLocation ?endpoint .
?entity_uri a :entity .
}}
{}
{}
}}
'''.format(filter_public_string, filter_entity_string)
query_launcher = SparqlQueryLauncher(self.app, self.session)
header, results = query_launcher.process_query(
self.prefix_query(query))
self.graphs = []
self.endpoints = []
for res in results:
if not graphs or res["graph"] in graphs:
self.graphs.append(res["graph"])
# If local triplestore url is not accessible by federetad query engine
if res["endpoint"] == self.settings.get(
'triplestore',
'endpoint') and self.local_endpoint_f is not None:
endpoint = self.local_endpoint_f
else:
endpoint = res["endpoint"]
if not endpoints or endpoint in endpoints:
self.endpoints.append(endpoint)
self.endpoints = Utils.unique(self.endpoints)
self.federated = len(self.endpoints) > 1
def get_file_name(self):
"""Get a random file name
Returns
-------
string
file name
"""
return Utils.get_random_string(10)
def replace_variables_in_triples(self, var_to_replace):
"""Replace variables in triples
Parameters
----------
var_to_replace : list of tuples
var to replace in triples
"""
for tpl_var in var_to_replace:
var_source = tpl_var[0]
var_target = tpl_var[1]
for i, triple_dict in enumerate(self.triples):
for key, value in triple_dict.items():
if key != "optional":
self.triples[i][key] = value.replace(var_source, var_target)
for i, select in enumerate(self.selects):
self.selects[i] = select.replace(var_source, var_target)
for i, filtr in enumerate(self.filters):
self.filters[i] = filtr.replace(var_source, var_target)
# uniq lists
self.triples = Utils.unique(self.triples)
self.selects = Utils.unique(self.selects)
def get_checked_asked_graphs(self, asked_graphs):
"""Check if asked graphs are present in public and private graphs
Parameters
----------
asked_graphs : list
list of graphs asked by the user
Returns
-------
list
list of graphs asked by the user, in the public and private graphs
"""
return Utils.intersect(asked_graphs, self.private_graphs + self.public_graphs)
def persist_user_admin(self, inputs):
"""Persist a new user (admin action)
Parameters
----------
inputs : User input
The new user info
Returns
-------
dict
The new user
"""
inputs["password"] = Utils.get_random_string(8)
return self.persist_user(inputs, return_password=True)
def update_users_table(self):
"""Add the quota col on the users table
Update the users table for the instance who don't have this column
"""
default_quota = Utils.humansize_to_bytes(self.settings.get("askomics", "quota"))
query = '''
ALTER TABLE users
ADD quota INTEGER NOT NULL DEFAULT {}
'''.format(default_quota)
try:
self.execute_sql_query(query)
except Exception:
pass
def create_reset_token(self, username, old_token=False):
"""Create a reset token"""
if old_token:
old_timestamp = int(time.time()) - 14400
token = "{}:{}".format(old_timestamp, Utils.get_random_string(20))
database = Database(self.app, self.session)
query = """
UPDATE users
SET reset_token=?
WHERE username=?
"""
database.execute_sql_query(query, (token, username))
return token
else:
auth = LocalAuth(self.app, self.session)
return auth.create_reset_token(username)
def __init__(self, app, session, result_info, force_no_db=False):
"""init object
Parameters
----------
app : Flask
flask app
session :
AskOmics session, contain the user
result_info : dict
Result file info
"""
Params.__init__(self, app, session)
if "user" in self.session:
self.result_path = "{}/{}_{}/results".format(
self.settings.get("askomics", "data_directory"),
self.session['user']['id'], self.session['user']['username'])
if "id" in result_info and not force_no_db:
self.id = result_info["id"]
if not self.set_info_from_db_with_id():
return None
else:
self.id = result_info["id"] if "id" in result_info else None
self.graph_state = result_info[
"graph_state"] if "graph_state" in result_info else None
self.graphs = result_info[
"graphs"] if "graphs" in result_info else []
self.endpoints = result_info[
"endpoints"] if "endpoints" in result_info else []
self.sparql_query = result_info[
"sparql_query"] if "sparql_query" in result_info else None
self.celery_id = result_info[
"celery_id"] if "celery_id" in result_info else None
self.file_name = result_info[
"file_name"] if "file_name" in result_info else Utils.get_random_string(
10)
self.file_path = "{}/{}".format(self.result_path, self.file_name)
self.start = None
self.end = None
self.nrows = 0
self.has_form_attr = False
self.template = False
self.form = False
def set_quota(self, quota, username):
"""Set a new quota to a user
Parameters
----------
quota : int
New quota
username : string
The concerned username
"""
database = Database(self.app, self.session)
query = '''
UPDATE users
SET quota=?
WHERE username=?
'''
database.execute_sql_query(query, (Utils.humansize_to_bytes(quota), username))
def rdfize(self, string):
"""Rdfize a string
Return the literal is string is an url, else,
prefix it with askomics prefix
Parameters
----------
string : string
Term to rdfize
Returns
-------
rdflib.???
Rdfized term
"""
if Utils.is_valid_url(string):
return rdflib.URIRef(string)
else:
return self.askomics_prefix[self.format_uri(string)]
def get_uri_filter_value(self, value):
"""Get full uri from a filter value (curie or uri)
:xxx --> :xxx
uniprot:xxx --> <http://purl.uniprot.org/core/xxx>
http://example.org/xxx --> <http://example.org/xxx>
xxx --> xxx is not a valid URI or CURIE
bla:xxx --> bla: is not a known prefix
Parameters
----------
value : string
Input filter
Returns
-------
string
corresponding uri
Raises
------
ValueError
Invalid URI or CURIE return a value error
"""
if Utils().is_url(value):
return "<{}>".format(value)
elif ":" in value:
prefix, val = value.split(":")
if prefix:
prefix_manager = PrefixManager(self.app, self.session)
namespace = prefix_manager.get_namespace(prefix)
if namespace:
return "<{}{}>".format(namespace, val)
else:
raise ValueError(
"{}: is not a known prefix".format(prefix))
else:
return value
raise ValueError("{} is not a valid URI or CURIE".format(value))
def download_datasets(self, datasets_id):
"""Download galaxy datasets into AskOmics
Parameters
----------
datasets_id : list
List of Galaxy datasets id
"""
galaxy_instance = galaxy.GalaxyInstance(self.url, self.apikey)
files_handler = FilesHandler(self.app, self.session)
for dataset_id in datasets_id:
dataset = galaxy_instance.datasets.show_dataset(dataset_id)
file_name = Utils.get_random_string(10)
path = "{}/{}".format(self.upload_path, file_name)
filetype = dataset["file_ext"]
size = dataset["file_size"]
galaxy_instance.datasets.download_dataset(
dataset_id, file_path=path, use_default_filename=False)
files_handler.store_file_info_in_db(dataset["name"], filetype,
file_name, size)
def integrate(self, public=False):
"""Integrate the file into the triplestore
Parameters
----------
public : bool, optional
Integrate in private or public graph
"""
sparql = SparqlQueryLauncher(self.app, self.session)
self.public = public
method = self.settings.get('triplestore', 'upload_method')
# insert metadata
sparql.insert_data(self.get_metadata(), self.file_graph, metadata=True)
if method == "load":
# cp file into ttl dir
tmp_file_name = 'tmp_{}_{}.ttl'.format(
Utils.get_random_string(5),
self.name,
)
temp_file_path = '{}/{}'.format(self.ttl_dir, tmp_file_name)
copyfile(self.path, temp_file_path)
# Load the chunk
sparql.load_data(tmp_file_name, self.file_graph, self.host_url)
# Remove tmp file
if not self.settings.getboolean('askomics', 'debug_ttl'):
os.remove(temp_file_path)
else:
with open(self.path) as ttl_file:
ttl_content = ttl_file.read()
sparql.insert_ttl_string(ttl_content, self.user_graph)
self.set_triples_number()
def update_pw_db(self, username, password):
"""Update a password in database
Parameters
----------
username : str
User username
password : str
New password
"""
database = Database(self.app, self.session)
salt = Utils.get_random_string(20)
salted_pw = self.settings.get('askomics', 'password_salt') + password + salt
sha512_pw = hashlib.sha512(salted_pw.encode('utf-8')).hexdigest()
query = '''
UPDATE users SET
password=?, salt=?
WHERE username=?
'''
database.execute_sql_query(query, (sha512_pw, salt, username))
def integrate(self, dataset_id=None):
"""Integrate the file into the triplestore"""
sparql = SparqlQueryLauncher(self.app, self.session)
# insert metadata
sparql.insert_data(self.get_metadata(), self.file_graph, metadata=True)
content_generator = self.generate_rdf_content()
# Insert content
chunk_number = 0
for _ in content_generator:
if self.graph_chunk.ntriple >= self.max_chunk_size:
if self.graph_chunk.percent and dataset_id:
self.update_percent_in_db(self.graph_chunk.percent,
dataset_id)
if self.method == 'load':
# write rdf into a tmpfile and load it
temp_file_name = 'tmp_{}_{}_chunk_{}.{}'.format(
Utils.get_random_string(5), self.name, chunk_number,
self.rdf_extention)
# Try to load data. if failure, wait 5 sec and retry 5 time
Utils.redo_if_failure(self.log, 5, 5, self.load_graph,
self.graph_chunk, temp_file_name)
else:
# Insert
# Try to insert data. if failure, wait 5 sec and retry 5 time
Utils.redo_if_failure(self.log, 5, 5, sparql.insert_data,
self.graph_chunk, self.file_graph)
chunk_number += 1
self.graph_chunk = RdfGraph(self.app, self.session)
# Load the last chunk
if self.graph_chunk.percent and dataset_id:
self.update_percent_in_db(100, dataset_id)
if self.method == 'load':
temp_file_name = 'tmp_{}_{}_chunk_{}.{}'.format(
Utils.get_random_string(5), self.name, chunk_number,
self.rdf_extention)
# Try to load data. if failure, wait 5 sec and retry 5 time
Utils.redo_if_failure(self.log, 5, 5, self.load_graph,
self.graph_chunk, temp_file_name)
else:
# Insert
# Try to insert data. if failure, wait 5 sec and retry 5 time
Utils.redo_if_failure(self.log, 5, 5, sparql.insert_data,
self.graph_chunk, self.file_graph)
# Content is inserted, now insert abstraction and domain_knowledge
self.set_rdf_abstraction_domain_knowledge()
if self.method == 'load':
temp_file_name = 'tmp_{}_{}_abstraction_domain_knowledge.{}'.format(
Utils.get_random_string(5), self.name, self.rdf_extention)
self.load_graph(self.graph_abstraction_dk, temp_file_name)
else:
# Insert
sparql.insert_data(self.graph_abstraction_dk, self.file_graph)
self.set_triples_number()
def build_query_from_json(self,
json_query,
preview=False,
for_editor=False):
"""Build a sparql query for the json dict of the query builder
Parameters
----------
json_query : dict
The json query from the query builder
Returns
-------
str
SPARQL query
"""
entities = []
attributes = {}
linked_attributes = []
self.selects = []
triples_relations = []
triples_attributes = []
values = []
filters = []
start_end = []
strands = []
var_to_replace = []
# Browse node to get graphs
for node in json_query["nodes"]:
if not node["suggested"]:
entities.append(node["uri"])
self.set_graphs_and_endpoints(entities=entities)
# self.log.debug(json_query)
# Browse links (relations)
for link in json_query["links"]:
if not link["suggested"]:
source = self.format_sparql_variable("{}{}_uri".format(
link["source"]["label"], link["source"]["id"]))
target = self.format_sparql_variable("{}{}_uri".format(
link["target"]["label"], link["target"]["id"]))
# Position
if link["uri"] in ('included_in', 'overlap_with'):
common_block = self.format_sparql_variable(
"block_{}_{}".format(link["source"]["id"],
link["target"]["id"]))
# Get start & end sparql variables
for attr in json_query["attr"]:
if not attr["faldo"]:
continue
if attr["nodeId"] == link["source"]["id"]:
if attr["faldo"].endswith("faldoStart"):
start_end.append(attr["id"])
start_1 = self.format_sparql_variable(
"{}{}_{}".format(attr["entityLabel"],
attr["nodeId"],
attr["label"]))
if attr["faldo"].endswith("faldoEnd"):
start_end.append(attr["id"])
end_1 = self.format_sparql_variable(
"{}{}_{}".format(attr["entityLabel"],
attr["nodeId"],
attr["label"]))
if attr["faldo"].endswith("faldoStrand"):
strand_1 = self.format_sparql_variable(
"{}{}_{}_faldoStrand".format(
attr["entityLabel"], attr["nodeId"],
attr["label"]))
strands.append(attr["id"])
if attr["nodeId"] == link["target"]["id"]:
if attr["faldo"].endswith("faldoStart"):
start_end.append(attr["id"])
start_2 = self.format_sparql_variable(
"{}{}_{}".format(attr["entityLabel"],
attr["nodeId"],
attr["label"]))
if attr["faldo"].endswith("faldoEnd"):
start_end.append(attr["id"])
end_2 = self.format_sparql_variable(
"{}{}_{}".format(attr["entityLabel"],
attr["nodeId"],
attr["label"]))
if attr["faldo"].endswith("faldoStrand"):
strand_2 = self.format_sparql_variable(
"{}{}_{}_faldoStrand".format(
attr["entityLabel"], attr["nodeId"],
attr["label"]))
strands.append(attr["id"])
triples_relations.append({
"subject":
source,
"predicate":
"askomics:{}".format("includeInReference" if
link["sameRef"] else "includeIn"),
"object":
common_block,
"optional":
False
})
triples_relations.append({
"subject":
target,
"predicate":
"askomics:{}".format("includeInReference" if
link["sameRef"] else "includeIn"),
"object":
common_block,
"optional":
False
})
if link["sameStrand"]:
var_to_replace.append((strand_1, strand_2))
else:
strands = []
equal_sign = "" if link["strict"] else "="
if link["uri"] == "included_in":
filters.append(
"FILTER ({start1} >{equalsign} {start2} && {end1} <{equalsign} {end2}) ."
.format(start1=start_1,
start2=start_2,
end1=end_1,
end2=end_2,
equalsign=equal_sign))
elif link["uri"] == "overlap_with":
filters.append(
"FILTER (({start2} >{equalsign} {start1} && {start2} <{equalsign} {end1}) || ({end2} >{equalsign} {start1} && {end2} <{equalsign} {end1}))"
.format(start1=start_1,
start2=start_2,
end1=end_1,
end2=end_2,
equalsign=equal_sign))
# Classic relation
else:
relation = "<{}>".format(link["uri"])
triples_relations.append({
"subject": source,
"predicate": relation,
"object": target,
"optional": False
})
# Store linked attributes
for attribute in json_query["attr"]:
attributes[attribute["id"]] = {
"label": attribute["label"],
"entity_label": attribute["entityLabel"],
"entity_id": attribute["nodeId"]
}
if attribute["linked"]:
linked_attributes.extend(
(attribute["id"], attribute["linkedWith"]))
# Browse attributes
for attribute in json_query["attr"]:
# URI ---
if attribute["type"] == "uri":
subject = self.format_sparql_variable("{}{}_uri".format(
attribute["entityLabel"], attribute["nodeId"]))
predicate = attribute["uri"]
obj = "<{}>".format(attribute["entityUri"])
if not self.is_bnode(attribute["entityUri"],
json_query["nodes"]):
triples_attributes.append({
"subject": subject,
"predicate": predicate,
"object": obj,
"optional": False
})
if attribute["visible"]:
self.selects.append(subject)
# filters/values
if attribute["filterValue"] != "" and not attribute["linked"]:
filter_value = self.get_uri_filter_value(
attribute["filterValue"])
if attribute["filterType"] == "regexp":
negative_sign = ""
if attribute["negative"]:
negative_sign = "!"
filters.append("FILTER ({}regex({}, {}, 'i'))".format(
negative_sign, subject, filter_value))
elif attribute["filterType"] == "exact":
if attribute["negative"]:
filters.append("FILTER (str({}) != {}) .".format(
subject, filter_value))
else:
values.append("VALUES {} {{ {} }} .".format(
subject, filter_value))
if attribute["linked"]:
var_2 = self.format_sparql_variable("{}{}_uri".format(
attributes[attribute["linkedWith"]]["entity_label"],
attributes[attribute["linkedWith"]]["entity_id"]))
var_to_replace.append((subject, var_2))
# Text
if attribute["type"] == "text":
if attribute["visible"] or attribute[
"filterValue"] != "" or attribute[
"id"] in linked_attributes:
subject = self.format_sparql_variable("{}{}_uri".format(
attribute["entityLabel"], attribute["nodeId"]))
if attribute["uri"] == "rdfs:label":
predicate = attribute["uri"]
else:
predicate = "<{}>".format(attribute["uri"])
obj = self.format_sparql_variable("{}{}_{}".format(
attribute["entityLabel"], attribute["nodeId"],
attribute["label"]))
triples_attributes.append({
"subject":
subject,
"predicate":
predicate,
"object":
obj,
"optional":
True if attribute["optional"] else False
})
if attribute["visible"]:
self.selects.append(obj)
# filters/values
if attribute["filterValue"] != "" and not attribute[
"optional"] and not attribute["linked"]:
if attribute["filterType"] == "regexp":
negative_sign = ""
if attribute["negative"]:
negative_sign = "!"
filters.append(
"FILTER ({}regex({}, '{}', 'i'))".format(
negative_sign, obj, attribute["filterValue"]))
elif attribute["filterType"] == "exact":
if attribute["negative"]:
filters.append("FILTER (str({}) != '{}') .".format(
obj, attribute["filterValue"]))
else:
values.append(
"VALUES {} {{ '{}'^^xsd:string }} .".format(
obj, attribute["filterValue"]))
if attribute["linked"]:
var_2 = self.format_sparql_variable("{}{}_{}".format(
attributes[attribute["linkedWith"]]["entity_label"],
attributes[attribute["linkedWith"]]["entity_id"],
attributes[attribute["linkedWith"]]["label"]))
var_to_replace.append((obj, var_2))
# Numeric
if attribute["type"] == "decimal":
if attribute["visible"] or attribute[
"filterValue"] != "" or attribute[
"id"] in start_end or attribute[
"id"] in linked_attributes:
subject = self.format_sparql_variable("{}{}_uri".format(
attribute["entityLabel"], attribute["nodeId"]))
if attribute["faldo"]:
predicate = "faldo:location/faldo:{}/faldo:position".format(
"begin" if attribute["faldo"].
endswith("faldoStart") else "end")
else:
predicate = "<{}>".format(attribute["uri"])
obj = self.format_sparql_variable("{}{}_{}".format(
attribute["entityLabel"], attribute["nodeId"],
attribute["label"]))
triples_attributes.append({
"subject":
subject,
"predicate":
predicate,
"object":
obj,
"optional":
True if attribute["optional"] else False
})
if attribute["visible"]:
self.selects.append(obj)
# filters
if attribute["filterValue"] != "" and not attribute[
"optional"] and not attribute["linked"]:
if attribute['filterSign'] == "=":
values.append("VALUES {} {{ {} }} .".format(
obj, attribute["filterValue"]))
else:
filter_string = "FILTER ( {} {} {} ) .".format(
obj, attribute["filterSign"],
attribute["filterValue"])
filters.append(filter_string)
if attribute["linked"]:
var_2 = self.format_sparql_variable("{}{}_{}".format(
attributes[attribute["linkedWith"]]["entity_label"],
attributes[attribute["linkedWith"]]["entity_id"],
attributes[attribute["linkedWith"]]["label"]))
var_to_replace.append((obj, var_2))
# Category
if attribute["type"] == "category":
if attribute["visible"] or attribute[
"filterSelectedValues"] != [] or attribute[
"id"] in strands or attribute[
"id"] in linked_attributes:
node_uri = self.format_sparql_variable("{}{}_uri".format(
attribute["entityLabel"], attribute["nodeId"]))
category_value_uri = self.format_sparql_variable(
"{}{}_{}Category".format(attribute["entityLabel"],
attribute["nodeId"],
attribute["label"]))
category_label = self.format_sparql_variable(
"{}{}_{}".format(attribute["entityLabel"],
attribute["nodeId"],
attribute["label"]))
faldo_strand = self.format_sparql_variable(
"{}{}_{}_faldoStrand".format(attribute["entityLabel"],
attribute["nodeId"],
attribute["label"]))
if attribute["faldo"] and attribute["faldo"].endswith(
"faldoReference"):
category_name = 'faldo:location/faldo:begin/faldo:reference'
triples_attributes.append({
"subject":
node_uri,
"predicate":
category_name,
"object":
category_value_uri,
"optional":
True if attribute["optional"] else False
})
if attribute["visible"]:
triples_attributes.append({
"subject":
category_value_uri,
"predicate":
"rdfs:label",
"object":
category_label,
"optional":
True if attribute["optional"] else False
})
elif attribute["faldo"] and attribute["faldo"].endswith(
"faldoStrand"):
category_name = 'faldo:location/faldo:begin/rdf:type'
triples_attributes.append({
"subject":
node_uri,
"predicate":
category_name,
"object":
category_value_uri,
"optional":
True if attribute["optional"] else False
})
triples_attributes.append({
"subject":
faldo_strand,
"predicate":
"a",
"object":
category_value_uri,
"optional":
True if attribute["optional"] else False
})
if attribute["visible"]:
triples_attributes.append({
"subject": faldo_strand,
"predicate": "rdfs:label",
"object": category_label,
"optional": False
})
values.append(
"VALUES {} {{ faldo:ReverseStrandPosition faldo:ForwardStrandPosition }} ."
.format(category_value_uri))
else:
category_name = "<{}>".format(attribute["uri"])
triples_attributes.append({
"subject":
node_uri,
"predicate":
category_name,
"object":
category_value_uri,
"optional":
True if attribute["optional"] else False
})
if attribute["visible"]:
triples_attributes.append({
"subject":
category_value_uri,
"predicate":
"rdfs:label",
"object":
category_label,
"optional":
True if attribute["optional"] else False
})
if attribute["visible"]:
self.selects.append(category_label)
# values
if attribute["filterSelectedValues"] != [] and not attribute[
"optional"] and not attribute["linked"]:
uri_val_list = []
for value in attribute["filterSelectedValues"]:
if attribute["faldo"] and attribute["faldo"].endswith(
"faldoStrand"):
value_var = faldo_strand
uri_val_list.append("<{}>".format(value))
else:
value_var = category_value_uri
uri_val_list.append("<{}>".format(value))
if uri_val_list:
values.append("VALUES {} {{ {} }}".format(
value_var, ' '.join(uri_val_list)))
if attribute["linked"]:
var_2 = self.format_sparql_variable(
"{}{}_{}Category".format(
attributes[
attribute["linkedWith"]]["entity_label"],
attributes[attribute["linkedWith"]]["entity_id"],
attributes[attribute["linkedWith"]]["label"]))
var_to_replace.append((category_value_uri, var_2))
from_string = self.get_froms_from_graphs(self.graphs)
federated_from_string = self.get_federated_froms_from_graphs(
self.graphs)
endpoints_string = self.get_endpoints_string()
# Linked attributes: replace SPARQL variable target by source
for tpl_var in var_to_replace:
var_source = tpl_var[0]
var_target = tpl_var[1]
for i, triple_dict in enumerate(triples_relations):
for key, value in triple_dict.items():
if key != "optional":
triples_relations[i][key] = value.replace(
var_source, var_target)
for i, triple_dict in enumerate(triples_attributes):
for key, value in triple_dict.items():
if key != "optional":
triples_attributes[i][key] = value.replace(
var_source, var_target)
for i, select in enumerate(self.selects):
self.selects[i] = select.replace(var_source, var_target)
for i, filtr in enumerate(filters):
filters[i] = filtr.replace(var_source, var_target)
# uniq lists
triples_relations = Utils.unique(triples_relations)
triples_attributes = Utils.unique(triples_attributes)
self.selects = Utils.unique(self.selects)
# Write the query
# query is for editor (no froms, no federated)
if for_editor:
query = """
SELECT DISTINCT {selects}
WHERE {{
{relations}
{attributes}
{filters}
{values}
}}
""".format(selects=' '.join(self.selects),
relations='\n '.join([
self.triple_dict_to_string(triple_dict)
for triple_dict in triples_relations
]),
attributes='\n '.join([
self.triple_dict_to_string(triple_dict)
for triple_dict in triples_attributes
]),
filters='\n '.join(filters),
values='\n '.join(values))
# Query is federated, add federated lines @federate & @from)
elif self.federated:
query = """
{endpoints}
{federated}
SELECT DISTINCT {selects}
WHERE {{
{relations}
{attributes}
{filters}
{values}
}}
""".format(endpoints=endpoints_string,
federated=federated_from_string,
selects=' '.join(self.selects),
relations='\n '.join([
self.triple_dict_to_string(triple_dict)
for triple_dict in triples_relations
]),
attributes='\n '.join([
self.triple_dict_to_string(triple_dict)
for triple_dict in triples_attributes
]),
filters='\n '.join(filters),
values='\n '.join(values))
# Query on the local endpoint (add froms)
elif self.endpoints == [self.local_endpoint_f]:
query = """
SELECT DISTINCT {selects}
{froms}
WHERE {{
{relations}
{attributes}
{filters}
{values}
}}
""".format(selects=' '.join(self.selects),
froms=from_string,
relations='\n '.join([
self.triple_dict_to_string(triple_dict)
for triple_dict in triples_relations
]),
attributes='\n '.join([
self.triple_dict_to_string(triple_dict)
for triple_dict in triples_attributes
]),
filters='\n '.join(filters),
values='\n '.join(values))
# Query an external endpoint (no froms)
else:
query = """
SELECT DISTINCT {selects}
WHERE {{
{relations}
{attributes}
{filters}
{values}
}}
""".format(selects=' '.join(self.selects),
relations='\n '.join([
self.triple_dict_to_string(triple_dict)
for triple_dict in triples_relations
]),
attributes='\n '.join([
self.triple_dict_to_string(triple_dict)
for triple_dict in triples_attributes
]),
filters='\n '.join(filters),
values='\n '.join(values))
if preview:
query += "\nLIMIT {}".format(
self.settings.getint('triplestore', 'preview_limit'))
return self.prefix_query(textwrap.dedent(query))
def build_query_from_json(self, preview=False, for_editor=False):
"""Build a sparql query for the json dict of the query builder
Parameters
----------
preview : bool, optional
Build a preview query (with LIMIT)
for_editor : bool, optional
Remove FROMS and @federate
"""
entities = []
attributes = {}
linked_attributes = []
self.selects = []
self.triples = []
self.triples_blocks = []
self.values = []
self.filters = []
start_end = []
strands = []
var_to_replace = []
# Browse attributes to get entities
for attr in self.json["attr"]:
if attr["visible"]:
entities = entities + attr["entityUris"]
entities = list(set(entities)) # uniq list
# Set graphs in function of entities needed
self.set_graphs_and_endpoints(entities=entities)
# self.log.debug(self.json)
# Browse links (relations)
for link in self.json["links"]:
if not link["suggested"]:
# if link is special, replace the special node variable with its real node
if link["type"] == "specialLink":
special_node = link["source"] if link["source"]["type"] in ("unionNode", "minusNode") else link["target"]
real_node = link["target"] if link["source"]["type"] in ("unionNode", "minusNode") else link["source"]
var_to_replace.append((
self.format_sparql_variable("{}{}_uri".format(special_node["label"], special_node["id"])),
self.format_sparql_variable("{}{}_uri".format(real_node["label"], real_node["id"]))
))
continue
source = self.format_sparql_variable("{}{}_uri".format(link["source"]["label"], link["source"]["id"]))
target = self.format_sparql_variable("{}{}_uri".format(link["target"]["label"], link["target"]["id"]))
# Check if relation is in a block
block_id = None
sblock_id = None
pblock_ids = (None, None)
if link["source"]["specialNodeId"] or link["target"]["specialNodeId"]:
block_id = link["source"]["specialNodeId"]
sblock_id = link["source"]["specialNodeGroupId"] if link["source"]["specialNodeGroupId"] else link["target"]["specialNodeGroupId"]
pblock_ids = link["source"]["specialPreviousIds"]
# Position
if link["uri"] in ('included_in', 'overlap_with'):
# If source of target is a special node, replace the id with the id of the concerned node
source_id = link["source"]["id"]
target_id = link["target"]["id"]
if link["source"]["type"] in ("unionNode", "minusNode"):
source_id = self.get_source_of_special_node(link["source"]["id"])
if link["target"]["type"] in ("unionNode", "minusNode"):
target_id = self.get_source_of_special_node(link["target"]["id"])
common_block = self.format_sparql_variable("block_{}_{}".format(link["source"]["id"], link["target"]["id"]))
# Get start & end sparql variables
for attr in self.json["attr"]:
if not attr["faldo"]:
continue
if attr["nodeId"] == source_id:
if attr["faldo"].endswith("faldoStart"):
start_end.append(attr["id"])
start_1 = self.format_sparql_variable("{}{}_{}".format(attr["entityLabel"], attr["nodeId"], attr["label"]))
if attr["faldo"].endswith("faldoEnd"):
start_end.append(attr["id"])
end_1 = self.format_sparql_variable("{}{}_{}".format(attr["entityLabel"], attr["nodeId"], attr["label"]))
if attr["faldo"].endswith("faldoStrand"):
strand_1 = self.format_sparql_variable("{}{}_{}_faldoStrand".format(attr["entityLabel"], attr["nodeId"], attr["label"]))
strands.append(attr["id"])
if attr["nodeId"] == target_id:
if attr["faldo"].endswith("faldoStart"):
start_end.append(attr["id"])
start_2 = self.format_sparql_variable("{}{}_{}".format(attr["entityLabel"], attr["nodeId"], attr["label"]))
if attr["faldo"].endswith("faldoEnd"):
start_end.append(attr["id"])
end_2 = self.format_sparql_variable("{}{}_{}".format(attr["entityLabel"], attr["nodeId"], attr["label"]))
if attr["faldo"].endswith("faldoStrand"):
strand_2 = self.format_sparql_variable("{}{}_{}_faldoStrand".format(attr["entityLabel"], attr["nodeId"], attr["label"]))
strands.append(attr["id"])
self.store_triple({
"subject": source,
"predicate": "askomics:{}".format("includeInReference" if link["sameRef"] else "includeIn"),
"object": common_block,
"optional": False
}, block_id, sblock_id, pblock_ids)
self.store_triple({
"subject": target,
"predicate": "askomics:{}".format("includeInReference" if link["sameRef"] else "includeIn"),
"object": common_block,
"optional": False
}, block_id, sblock_id, pblock_ids)
if link["sameStrand"]:
var_to_replace.append((strand_1, strand_2))
else:
strands = []
equal_sign = "" if link["strict"] else "="
if link["uri"] == "included_in":
self.store_filter("FILTER ({start1} >{equalsign} {start2} && {end1} <{equalsign} {end2}) .".format(
start1=start_1,
start2=start_2,
end1=end_1,
end2=end_2,
equalsign=equal_sign
), block_id, sblock_id, pblock_ids)
elif link["uri"] == "overlap_with":
self.store_filter("FILTER (({start2} >{equalsign} {start1} && {start2} <{equalsign} {end1}) || ({end2} >{equalsign} {start1} && {end2} <{equalsign} {end1}))".format(
start1=start_1,
start2=start_2,
end1=end_1,
end2=end_2,
equalsign=equal_sign
), block_id, sblock_id, pblock_ids)
# Classic relation
else:
relation = "<{}>".format(link["uri"])
triple = {
"subject": source,
"predicate": relation,
"object": target,
"optional": False
}
self.store_triple(triple, block_id, sblock_id, pblock_ids)
# Store linked attributes
for attribute in self.json["attr"]:
attributes[attribute["id"]] = {
"label": attribute["label"],
"entity_label": attribute["entityLabel"],
"entity_id": attribute["nodeId"]
}
if attribute["linked"]:
linked_attributes.extend((attribute["id"], attribute["linkedWith"]))
# Browse attributes
for attribute in self.json["attr"]:
# Get blockid and sblockid of the attribute node
block_id, sblock_id, pblock_ids = self.get_block_ids(attribute["nodeId"])
# URI ---
if attribute["type"] == "uri":
subject = self.format_sparql_variable("{}{}_uri".format(attribute["entityLabel"], attribute["nodeId"]))
predicate = attribute["uri"]
obj = "<{}>".format(attribute["entityUris"][0])
if not self.is_bnode(attribute["entityUris"][0], self.json["nodes"]):
self.store_triple({
"subject": subject,
"predicate": predicate,
"object": obj,
"optional": False
}, block_id, sblock_id, pblock_ids)
if attribute["visible"]:
self.selects.append(subject)
# filters/values
if attribute["filterValue"] != "" and not attribute["linked"]:
filter_value = self.get_uri_filter_value(attribute["filterValue"])
if attribute["filterType"] == "regexp":
negative_sign = ""
if attribute["negative"]:
negative_sign = "!"
self.store_filter("FILTER ({}regex({}, {}, 'i'))".format(negative_sign, subject, filter_value), block_id, sblock_id, pblock_ids)
elif attribute["filterType"] == "exact":
if attribute["negative"]:
self.store_filter("FILTER (str({}) != {}) .".format(subject, filter_value), block_id, sblock_id, pblock_ids)
else:
self.store_value("VALUES {} {{ {} }} .".format(subject, filter_value), block_id, sblock_id, pblock_ids)
if attribute["linked"]:
var_2 = self.format_sparql_variable("{}{}_uri".format(
attributes[attribute["linkedWith"]]["entity_label"],
attributes[attribute["linkedWith"]]["entity_id"]
))
var_to_replace.append((subject, var_2))
if attribute["type"] == "boolean":
if attribute["visible"] or attribute["filterSelectedValues"] != [] or attribute["id"] in linked_attributes:
subject = self.format_sparql_variable("{}{}_uri".format(attribute["entityLabel"], attribute["nodeId"]))
predicate = "<{}>".format(attribute["uri"])
obj = self.format_sparql_variable("{}{}_{}".format(attribute["entityLabel"], attribute["humanNodeId"], attribute["label"]))
self.store_triple({
"subject": subject,
"predicate": predicate,
"object": obj,
"optional": True if attribute["optional"] else False
}, block_id, sblock_id, pblock_ids)
if attribute["visible"]:
self.selects.append(obj)
# values
if attribute["filterSelectedValues"] != [] and not attribute["optional"] and not attribute["linked"]:
uri_val_list = []
for value in attribute["filterSelectedValues"]:
if value == "true":
bool_value = "'true'^^xsd:boolean"
else:
bool_value = "'false'^^xsd:boolean"
value_var = obj
uri_val_list.append(bool_value)
if uri_val_list:
self.store_value("VALUES {} {{ {} }}".format(value_var, ' '.join(uri_val_list)), block_id, sblock_id, pblock_ids)
if attribute["linked"]:
var_2 = self.format_sparql_variable("{}{}_{}".format(
attributes[attribute["linkedWith"]]["entity_label"],
attributes[attribute["linkedWith"]]["entity_id"],
attributes[attribute["linkedWith"]]["label"]
))
var_to_replace.append((obj, var_2))
# Text
if attribute["type"] == "text":
if attribute["visible"] or attribute["filterValue"] != "" or attribute["id"] in linked_attributes:
subject = self.format_sparql_variable("{}{}_uri".format(attribute["entityLabel"], attribute["nodeId"]))
if attribute["uri"] == "rdfs:label":
predicate = attribute["uri"]
else:
predicate = "<{}>".format(attribute["uri"])
obj = self.format_sparql_variable("{}{}_{}".format(attribute["entityLabel"], attribute["humanNodeId"], attribute["label"]))
self.store_triple({
"subject": subject,
"predicate": predicate,
"object": obj,
"optional": True if attribute["optional"] else False
}, block_id, sblock_id, pblock_ids)
if attribute["visible"]:
self.selects.append(obj)
# filters/values
if attribute["filterValue"] != "" and not attribute["optional"] and not attribute["linked"]:
if attribute["filterType"] == "regexp":
negative_sign = ""
if attribute["negative"]:
negative_sign = "!"
self.store_filter("FILTER ({}regex({}, '{}', 'i'))".format(negative_sign, obj, attribute["filterValue"]), block_id, sblock_id, pblock_ids)
elif attribute["filterType"] == "exact":
if attribute["negative"]:
self.store_filter("FILTER (str({}) != '{}') .".format(obj, attribute["filterValue"]), block_id, sblock_id, pblock_ids)
else:
self.store_value("VALUES {} {{ '{}' }} .".format(obj, attribute["filterValue"]), block_id, sblock_id, pblock_ids)
if attribute["linked"]:
var_2 = self.format_sparql_variable("{}{}_{}".format(
attributes[attribute["linkedWith"]]["entity_label"],
attributes[attribute["linkedWith"]]["entity_id"],
attributes[attribute["linkedWith"]]["label"]
))
var_to_replace.append((obj, var_2))
# Numeric
if attribute["type"] == "decimal":
if attribute["visible"] or Utils.check_key_in_list_of_dict(attribute["filters"], "filterValue") or attribute["id"] in start_end or attribute["id"] in linked_attributes:
subject = self.format_sparql_variable("{}{}_uri".format(attribute["entityLabel"], attribute["nodeId"]))
if attribute["faldo"]:
predicate = "faldo:location/faldo:{}/faldo:position".format("begin" if attribute["faldo"].endswith("faldoStart") else "end")
else:
predicate = "<{}>".format(attribute["uri"])
obj = self.format_sparql_variable("{}{}_{}".format(attribute["entityLabel"], attribute["nodeId"], attribute["label"]))
self.store_triple({
"subject": subject,
"predicate": predicate,
"object": obj,
"optional": True if attribute["optional"] else False
}, block_id, sblock_id, pblock_ids)
if attribute["visible"]:
self.selects.append(obj)
# filters
for filtr in attribute["filters"]:
if filtr["filterValue"] != "" and not attribute["optional"] and not attribute["linked"]:
if filtr['filterSign'] == "=":
self.store_value("VALUES {} {{ {} }} .".format(obj, filtr["filterValue"]), block_id, sblock_id, pblock_ids)
else:
filter_string = "FILTER ( {} {} {} ) .".format(obj, filtr["filterSign"], filtr["filterValue"])
self.store_filter(filter_string, block_id, sblock_id, pblock_ids)
if attribute["linked"]:
var_2 = self.format_sparql_variable("{}{}_{}".format(
attributes[attribute["linkedWith"]]["entity_label"],
attributes[attribute["linkedWith"]]["entity_id"],
attributes[attribute["linkedWith"]]["label"]
))
var_to_replace.append((obj, var_2))
# Category
if attribute["type"] == "category":
if attribute["visible"] or attribute["filterSelectedValues"] != [] or attribute["id"] in strands or attribute["id"] in linked_attributes:
node_uri = self.format_sparql_variable("{}{}_uri".format(attribute["entityLabel"], attribute["nodeId"]))
category_value_uri = self.format_sparql_variable("{}{}_{}Category".format(attribute["entityLabel"], attribute["nodeId"], attribute["label"]))
category_label = self.format_sparql_variable("{}{}_{}".format(attribute["entityLabel"], attribute["humanNodeId"], attribute["label"]))
faldo_strand = self.format_sparql_variable("{}{}_{}_faldoStrand".format(attribute["entityLabel"], attribute["nodeId"], attribute["label"]))
if attribute["faldo"] and attribute["faldo"].endswith("faldoReference"):
category_name = 'faldo:location/faldo:begin/faldo:reference'
self.store_triple({
"subject": node_uri,
"predicate": category_name,
"object": category_value_uri,
"optional": True if attribute["optional"] else False
}, block_id, sblock_id, pblock_ids)
if attribute["visible"]:
self.store_triple({
"subject": category_value_uri,
"predicate": "rdfs:label",
"object": category_label,
"optional": True if attribute["optional"] else False
}, block_id, sblock_id, pblock_ids)
elif attribute["faldo"] and attribute["faldo"].endswith("faldoStrand"):
category_name = 'faldo:location/faldo:begin/rdf:type'
self.store_triple({
"subject": node_uri,
"predicate": category_name,
"object": category_value_uri,
"optional": True if attribute["optional"] else False
}, block_id, sblock_id, pblock_ids)
self.store_triple({
"subject": faldo_strand,
"predicate": "a",
"object": category_value_uri,
"optional": True if attribute["optional"] else False
}, block_id, sblock_id, pblock_ids)
if attribute["visible"]:
self.store_triple({
"subject": faldo_strand,
"predicate": "rdfs:label",
"object": category_label,
"optional": False
}, block_id, sblock_id, pblock_ids)
self.store_value("VALUES {} {{ faldo:ReverseStrandPosition faldo:ForwardStrandPosition }} .".format(category_value_uri), block_id, sblock_id, pblock_ids)
else:
category_name = "<{}>".format(attribute["uri"])
self.store_triple({
"subject": node_uri,
"predicate": category_name,
"object": category_value_uri,
"optional": True if attribute["optional"] else False
}, block_id, sblock_id, pblock_ids)
if attribute["visible"]:
self.store_triple({
"subject": category_value_uri,
"predicate": "rdfs:label",
"object": category_label,
"optional": True if attribute["optional"] else False
}, block_id, sblock_id, pblock_ids)
if attribute["visible"]:
self.selects.append(category_label)
# values
if attribute["filterSelectedValues"] != [] and not attribute["optional"] and not attribute["linked"]:
uri_val_list = []
for value in attribute["filterSelectedValues"]:
if attribute["faldo"] and attribute["faldo"].endswith("faldoStrand"):
value_var = faldo_strand
uri_val_list.append("<{}>".format(value))
else:
value_var = category_value_uri
uri_val_list.append("<{}>".format(value))
if uri_val_list:
self.store_value("VALUES {} {{ {} }}".format(value_var, ' '.join(uri_val_list)), block_id, sblock_id, pblock_ids)
if attribute["linked"]:
var_2 = self.format_sparql_variable("{}{}_{}Category".format(
attributes[attribute["linkedWith"]]["entity_label"],
attributes[attribute["linkedWith"]]["entity_id"],
attributes[attribute["linkedWith"]]["label"]
))
var_to_replace.append((category_value_uri, var_2))
from_string = self.get_froms_from_graphs(self.graphs)
federated_from_string = self.get_federated_froms_from_graphs(self.graphs)
endpoints_string = self.get_endpoints_string()
# Linked attributes: replace SPARQL variable target by source
self.replace_variables_in_blocks(var_to_replace)
self.replace_variables_in_triples(var_to_replace)
# Write the query
# query is for editor (no froms, no federated)
if for_editor:
query = """
SELECT DISTINCT {selects}
WHERE {{
{triples}
{blocks}
{filters}
{values}
}}
""".format(
selects=' '.join(self.selects),
triples='\n '.join([self.triple_dict_to_string(triple_dict) for triple_dict in self.triples]),
blocks='\n '.join([self.triple_block_to_string(triple_block) for triple_block in self.triples_blocks]),
filters='\n '.join(self.filters),
values='\n '.join(self.values))
# Query is federated, add federated lines @federate & @from)
elif self.federated:
query = """
{endpoints}
{federated}
SELECT DISTINCT {selects}
WHERE {{
{triples}
{blocks}
{filters}
{values}
}}
""".format(
endpoints=endpoints_string,
federated=federated_from_string,
selects=' '.join(self.selects),
triples='\n '.join([self.triple_dict_to_string(triple_dict) for triple_dict in self.triples]),
blocks='\n '.join([self.triple_block_to_string(triple_block) for triple_block in self.triples_blocks]),
filters='\n '.join(self.filters),
values='\n '.join(self.values)
)
# Query on the local endpoint (add froms)
elif self.endpoints == [self.local_endpoint_f]:
query = """
SELECT DISTINCT {selects}
{froms}
WHERE {{
{triples}
{blocks}
{filters}
{values}
}}
""".format(
selects=' '.join(self.selects),
froms=from_string,
triples='\n '.join([self.triple_dict_to_string(triple_dict) for triple_dict in self.triples]),
blocks='\n '.join([self.triple_block_to_string(triple_block) for triple_block in self.triples_blocks]),
filters='\n '.join(self.filters),
values='\n '.join(self.values))
# Query an external endpoint (no froms)
else:
query = """
SELECT DISTINCT {selects}
WHERE {{
{triples}
{blocks}
{filters}
{values}
}}
""".format(
selects=' '.join(self.selects),
triples='\n '.join([self.triple_dict_to_string(triple_dict) for triple_dict in self.triples]),
blocks='\n '.join([self.triple_block_to_string(triple_block) for triple_block in self.triples_blocks]),
filters='\n '.join(self.filters),
values='\n '.join(self.values))
if preview:
query += "\nLIMIT {}".format(self.settings.getint('triplestore', 'preview_limit'))
self.sparql = self.prefix_query(textwrap.dedent(query))
def persist_user(self, inputs, ldap_login=False, return_password=False):
"""
Persist user in the TS
Parameters
----------
inputs : dict
User infos
ldap_login : bool, optional
If True, user is ldap
Returns
-------
dict
The user
"""
database = Database(self.app, self.session)
# Check if user is the first. if yes, set him admin
if self.get_number_of_users() == 0:
admin = True
blocked = False
else:
admin = False
blocked = self.settings.getboolean('askomics', 'default_locked_account')
api_key = Utils.get_random_string(20) if "apikey" not in inputs else inputs["apikey"]
query = '''
INSERT INTO users VALUES(
NULL,
?,
?,
?,
?,
?,
?,
?,
?,
?,
?,
?,
?,
NULL
)
'''
salt = None
sha512_pw = None
email = None
fname = None
lname = None
if not ldap_login:
# Create a salt
salt = Utils.get_random_string(20) if "salt" not in inputs else inputs["salt"]
# Concat askomics_salt + user_password + salt
salted_pw = self.settings.get('askomics', 'password_salt') + inputs['password'] + salt
# hash
sha512_pw = hashlib.sha512(salted_pw.encode('utf8')).hexdigest()
email = inputs["email"]
fname = inputs["fname"]
lname = inputs["lname"]
# Store user in db
user_id = database.execute_sql_query(
query, (ldap_login, fname, lname, inputs['username'],
email, sha512_pw, salt, api_key, admin, blocked, Utils.humansize_to_bytes(self.settings.get("askomics", "quota")), int(time.time())), True)
user = {
'id': user_id,
'ldap': ldap_login,
'fname': fname,
'lname': lname,
'username': inputs['username'],
'email': email,
'admin': admin,
'blocked': blocked,
'quota': Utils.humansize_to_bytes(self.settings.get("askomics", "quota")),
'apikey': api_key,
'galaxy': None
}
if return_password and not ldap_login:
user["password"] = inputs["password"]
# Return user infos
return user
