Skip to content

Commit

Permalink
refactor test subpackage
Browse files Browse the repository at this point in the history
  • Loading branch information
@Remi-Gau
Remi-Gau committed Feb 20, 2024
1 parent e9b572c commit 059d30a
Show file tree
Hide file tree
Showing 2 changed files with 96 additions and 109 deletions.
2 changes: 1 addition & 1 deletion nidmresults/test/check_consistency.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -10,7 +10,7 @@

from rdflib import RDF

from nidmresults.objects.constants import *
from nidmresults.objects.constants import OWL

RELPATH = os.path.dirname(os.path.abspath(__file__))

Expand Down
203 changes: 95 additions & 108 deletions nidmresults/test/test_results_doc.py → nidmresults/test/utils.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -18,7 +18,12 @@
from rdflib.graph import Graph
from rdflib.namespace import RDF

from nidmresults.objects.constants_rdflib import *
from nidmresults.objects.constants_rdflib import (
DCT,
NIDM_IN_COORDINATE_SPACE,
PROV,
XSD,
)
from nidmresults.owl.owl_reader import OwlReader

# Append parent script directory to path
Expand All @@ -31,13 +36,10 @@ class TestResultDataModel:

def get_readable_name(self, owl, graph, item):
if isinstance(item, rdflib.term.Literal):
if item.datatype:
typeStr = graph.qname(item.datatype)
else:
typeStr = ""
typeStr = graph.qname(item.datatype) if item.datatype else ""
if typeStr:
typeStr = "(" + typeStr + ")"
name = "'" + item + "'" + typeStr
typeStr = f"({typeStr})"
name = f"'{item}'{typeStr}"
elif isinstance(item, rdflib.term.URIRef):
# Look for label
# name = graph.label(item)
Expand All @@ -54,30 +56,20 @@ def get_readable_name(self, owl, graph, item):
if m is not None:
name += " (i.e. " + owl.get_label(item).split(":")[1] + ")"
else:
name = "unsupported type: " + item
name = f"unsupported type: {item}"
return name

def get_alternatives(self, owl, graph, s=None, p=None, o=None):
found = ""

for s_in, p_in, o_in in graph.triples((s, p, o)):
if not o:
if not o_in.startswith(str(PROV)):
found += "; " + self.get_readable_name(
owl,
graph,
o_in,
)
if not p:
if not p_in.startswith(str(PROV)):
found += "; " + self.get_readable_name(owl, graph, p_in)
if not s:
if not s_in.startswith(str(PROV)):
found += "; " + self.get_readable_name(owl, graph, s_in)
if len(found) > 200:
found = "<many alternatives>"
else:
found = found[2:]
if not o and not o_in.startswith(str(PROV)):
found += f"; {self.get_readable_name(owl, graph, o_in)}"
if not p and not p_in.startswith(str(PROV)):
found += f"; {self.get_readable_name(owl, graph, p_in)}"
if not s and not s_in.startswith(str(PROV)):
found += f"; {self.get_readable_name(owl, graph, s_in)}"
found = "<many alternatives>" if len(found) > 200 else found[2:]
return found

# FIXME: Extend tests to more than one dataset (group analysis, ...)
Expand All @@ -92,7 +84,7 @@ def get_alternatives(self, owl, graph, s=None, p=None, o=None):
def setUp(self, parent_gt_dir=None):
self.my_execption = ""
self.gt_dir = parent_gt_dir
self.ex_graphs = dict()
self.ex_graphs = {}

def load_graph(self, ttl_name):
if ttl_name not in self.ex_graphs:
Expand Down Expand Up @@ -149,20 +141,16 @@ def print_results(self, res):
def successful_retreive(self, res, info_str=""):
"""Check if the results query 'res' contains a value for each field."""
if not res.bindings:
self.my_execption = info_str + """: Empty query results"""
self.my_execption = f"""{info_str}: Empty query results"""
return False
for idx, row in enumerate(res.bindings):
for row in res.bindings:
for key, val in sorted(row.items()):
logging.debug(f"{key}-->{val.decode()}")
if not val.decode():
self.my_execption += "\nMissing: \t %s" % (key)
return False
return True

# if not self.successful_retreive(self.spmexport.query(query),
# 'ContrastMap and ContrastStandardErrorMap'):
# raise Exception(self.my_execption)

def _replace_match(self, graph1, graph2, rdf_type):
"""Match classes of type 'rdf_type' across documents \
based on attributes."""
Expand Down Expand Up @@ -271,54 +259,53 @@ def _replace_match(self, graph1, graph2, rdf_type):
match_found = False
g2_matched = set()
for g2_term, match_index in list(g2_match.items()):
if max(g2_match.values()) >= min_matching:
if (
match_index == max(g2_match.values())
) and g2_term not in g2_matched:
# Found matching term
g2_matched.add(g2_term)

if not g1_term == g2_term:
g2_name = graph2.qname(g2_term).split(":")[-1]
new_id = g1_term + "_" + g2_name
logging.debug(
graph1.qname(g1_term)
+ " is matched to "
+ graph2.qname(g2_term)
+ " and replaced by "
+ graph2.qname(new_id)
+ " (match="
+ str(match_index)
+ ")"
)
if max(g2_match.values()) >= min_matching and (
(match_index == max(g2_match.values()))
and g2_term not in g2_matched
):
g2_matched.add(g2_term)

if not g1_term == g2_term:
g2_name = graph2.qname(g2_term).split(":")[-1]
new_id = g1_term + "_" + g2_name
logging.debug(
graph1.qname(g1_term)
+ " is matched to "
+ graph2.qname(g2_term)
+ " and replaced by "
+ graph2.qname(new_id)
+ " (match="
+ str(match_index)
+ ")"
)

for p, o in graph1.predicate_objects(g1_term):
graph1.remove((g1_term, p, o))
graph1.add((new_id, p, o))
for p, o in graph2.predicate_objects(g2_term):
graph2.remove((g2_term, p, o))
graph2.add((new_id, p, o))
for s, p in graph1.subject_predicates(g1_term):
graph1.remove((s, p, g1_term))
graph1.add((s, p, new_id))
for s, p in graph2.subject_predicates(g2_term):
graph2.remove((s, p, g2_term))
graph2.add((s, p, new_id))

g2_terms.remove(g2_term)
g2_terms.add(new_id)
else:
logging.debug(
graph1.qname(g1_term)
+ " is matched to "
+ graph2.qname(g2_term)
+ " (match="
+ str(match_index)
+ ")"
)
for p, o in graph1.predicate_objects(g1_term):
graph1.remove((g1_term, p, o))
graph1.add((new_id, p, o))
for p, o in graph2.predicate_objects(g2_term):
graph2.remove((g2_term, p, o))
graph2.add((new_id, p, o))
for s, p in graph1.subject_predicates(g1_term):
graph1.remove((s, p, g1_term))
graph1.add((s, p, new_id))
for s, p in graph2.subject_predicates(g2_term):
graph2.remove((s, p, g2_term))
graph2.add((s, p, new_id))

g2_terms.remove(g2_term)
g2_terms.add(new_id)
else:
logging.debug(
graph1.qname(g1_term)
+ " is matched to "
+ graph2.qname(g2_term)
+ " (match="
+ str(match_index)
+ ")"
)

match_found = True
break
match_found = True
break

if not match_found:
logging.debug("No match found for " + graph1.qname(g1_term))
Expand Down Expand Up @@ -362,9 +349,9 @@ def compare_full_graphs(
gt_graph, other_graph
)

in_both, in_gt, in_other = graph_diff(gt_graph, other_graph)
_, in_gt, in_other = graph_diff(gt_graph, other_graph)

exc_missing = list()
exc_missing = []

for s, p, o in in_gt:
# If there is a corresponding s,p check if
Expand All @@ -388,19 +375,18 @@ def compare_full_graphs(
)
)

exc_added = list()
exc_added = []
if not include:
for s, p, o in in_other:
if p not in to_ignore:
exc_added.append(
"\nAdded :\t '%s %s %s'"
% (
self.get_readable_name(owl, other_graph, s),
self.get_readable_name(owl, other_graph, p),
self.get_readable_name(owl, other_graph, o),
)
)

exc_added.extend(
"\nAdded :\t '%s %s %s'"
% (
self.get_readable_name(owl, other_graph, s),
self.get_readable_name(owl, other_graph, p),
self.get_readable_name(owl, other_graph, o),
)
for s, p, o in in_other
if p not in to_ignore
)
my_exception += "".join(sorted(exc_missing) + sorted(exc_added))

if raise_now and my_exception:
Expand Down Expand Up @@ -432,22 +418,23 @@ def _same_json_or_float(self, o, o_other):
same_json_array = False

numeric_types = [XSD.float, XSD.double, XSD.long, XSD.int]
if o.datatype in numeric_types:
if o_other.datatype in numeric_types:
# If both are zero but of different type isclose returns
# false
if o.value == 0 and o_other.value == 0:
close_float = True

# Avoid None
if o.value and o_other.value:
close_float = np.isclose(o.value, o_other.value)

if o.datatype in [XSD.string, None]:
if o_other.datatype in [XSD.string, None]:
if o.value == o_other.value:
same_str = True

if (
o.datatype in numeric_types
and o_other.datatype in numeric_types
):
if o.value == 0 and o_other.value == 0:
close_float = True

# Avoid None
if o.value and o_other.value:
close_float = np.isclose(o.value, o_other.value)

if (
o.datatype in [XSD.string, None]
and o_other.datatype in [XSD.string, None]
and o.value == o_other.value
):
same_str = True
return (same_json_array, close_float, same_str)


Expand Down

0 comments on commit 059d30a

Please sign in to comment.