test for issue #725: canonicalization collapses BNodes

1 files changed, 122 insertions, 12 deletions

diff --git a/test/test_canonicalization.py b/test/test_canonicalization.py
index b64059e7..2cdeda2c 100644
--- a/test/test_canonicalization.py
+++ b/test/test_canonicalization.py
@@ -1,3 +1,4 @@
+from collections import Counter
 from rdflib import Graph, RDF, BNode, URIRef, Namespace, ConjunctiveGraph, Literal
 from rdflib.compare import to_isomorphic, to_canonical_graph
 from rdflib.plugins.memory import IOMemory
@@ -220,28 +221,28 @@ def test_issue494_collapsing_bnodes():
          RDF['Statement']),
     ]
 
-    print('graph length: %d, nodes: %d' % (len(g), len(g.all_nodes())))
-    print('triple_bnode degrees:')
-    for triple_bnode in g.subjects(RDF['type'], RDF['Statement']):
-        print(len(list(g.triples([triple_bnode, None, None]))))
-    print('all node degrees:')
+    # print('graph length: %d, nodes: %d' % (len(g), len(g.all_nodes())))
+    # print('triple_bnode degrees:')
+    # for triple_bnode in g.subjects(RDF['type'], RDF['Statement']):
+    #     print(len(list(g.triples([triple_bnode, None, None]))))
+    # print('all node degrees:')
     g_node_degs = sorted([
         len(list(g.triples([node, None, None])))
         for node in g.all_nodes()
     ], reverse=True)
-    print(g_node_degs)
+    # print(g_node_degs)
 
     cg = to_canonical_graph(g)
-    print('graph length: %d, nodes: %d' % (len(cg), len(cg.all_nodes())))
-    print('triple_bnode degrees:')
-    for triple_bnode in cg.subjects(RDF['type'], RDF['Statement']):
-        print(len(list(cg.triples([triple_bnode, None, None]))))
-    print('all node degrees:')
+    # print('graph length: %d, nodes: %d' % (len(cg), len(cg.all_nodes())))
+    # print('triple_bnode degrees:')
+    # for triple_bnode in cg.subjects(RDF['type'], RDF['Statement']):
+    #     print(len(list(cg.triples([triple_bnode, None, None]))))
+    # print('all node degrees:')
     cg_node_degs = sorted([
         len(list(cg.triples([node, None, None])))
         for node in cg.all_nodes()
     ], reverse=True)
-    print(cg_node_degs)
+    # print(cg_node_degs)
 
     assert len(g) == len(cg), \
         'canonicalization changed number of triples in graph'
@@ -253,6 +254,24 @@ def test_issue494_collapsing_bnodes():
     assert g_node_degs == cg_node_degs, \
         'canonicalization changed node degrees'
 
+    # counter for subject, predicate and object nodes
+    g_pos_counts = Counter(), Counter(), Counter()
+    for t in g:
+        for i, node in enumerate(t):
+            g_pos_counts[i][t] += 1
+    g_count_signature = [sorted(c.values()) for c in g_pos_counts]
+
+    cg = to_canonical_graph(g)
+    cg_pos_counts = Counter(), Counter(), Counter()
+    for t in cg:
+        for i, node in enumerate(t):
+            cg_pos_counts[i][t] += 1
+    cg_count_signature = [sorted(c.values()) for c in cg_pos_counts]
+
+    assert g_count_signature == cg_count_signature, \
+        'canonicalization changed node position counts'
+
+
 def test_issue682_signing_named_graphs():
     ns = Namespace("http://love.com#")
 
@@ -282,3 +301,94 @@ def test_issue682_signing_named_graphs():
     assert len(ig) == len(g)
     assert len(igmary) < len(ig)
     assert ig.graph_digest() != igmary.graph_digest()
+
+
+def test_issue725_collapsing_bnodes():
+    g = Graph()
+    g += [
+        (BNode('N0a76d42406b84fe4b8029d0a7fa04244'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#object'),
+         BNode('v2')),
+        (BNode('N0a76d42406b84fe4b8029d0a7fa04244'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#predicate'),
+         BNode('v0')),
+        (BNode('N0a76d42406b84fe4b8029d0a7fa04244'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#subject'),
+         URIRef(u'urn:gp_learner:fixed_var:target')),
+        (BNode('N0a76d42406b84fe4b8029d0a7fa04244'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#type'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#Statement')),
+        (BNode('N2f62af5936b94a8eb4b1e4bfa8e11d95'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#object'),
+         BNode('v1')),
+        (BNode('N2f62af5936b94a8eb4b1e4bfa8e11d95'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#predicate'),
+         BNode('v0')),
+        (BNode('N2f62af5936b94a8eb4b1e4bfa8e11d95'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#subject'),
+         URIRef(u'urn:gp_learner:fixed_var:target')),
+        (BNode('N2f62af5936b94a8eb4b1e4bfa8e11d95'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#type'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#Statement')),
+        (BNode('N5ae541f93e1d4e5880450b1bdceb6404'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#object'),
+         BNode('v5')),
+        (BNode('N5ae541f93e1d4e5880450b1bdceb6404'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#predicate'),
+         BNode('v4')),
+        (BNode('N5ae541f93e1d4e5880450b1bdceb6404'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#subject'),
+         URIRef(u'urn:gp_learner:fixed_var:target')),
+        (BNode('N5ae541f93e1d4e5880450b1bdceb6404'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#type'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#Statement')),
+        (BNode('N86ac7ca781f546ae939b8963895f672e'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#object'),
+         URIRef(u'urn:gp_learner:fixed_var:source')),
+        (BNode('N86ac7ca781f546ae939b8963895f672e'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#predicate'),
+         BNode('v0')),
+        (BNode('N86ac7ca781f546ae939b8963895f672e'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#subject'),
+         URIRef(u'urn:gp_learner:fixed_var:target')),
+        (BNode('N86ac7ca781f546ae939b8963895f672e'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#type'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#Statement')),
+        (BNode('Nac82b883ca3849b5ab6820b7ac15e490'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#object'),
+         BNode('v1')),
+        (BNode('Nac82b883ca3849b5ab6820b7ac15e490'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#predicate'),
+         BNode('v3')),
+        (BNode('Nac82b883ca3849b5ab6820b7ac15e490'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#subject'),
+         URIRef(u'urn:gp_learner:fixed_var:target')),
+        (BNode('Nac82b883ca3849b5ab6820b7ac15e490'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#type'),
+         URIRef(u'http://www.w3.org/1999/02/22-rdf-syntax-ns#Statement'))
+    ]
+
+    cg = to_canonical_graph(g)
+    assert len(g) == len(cg), \
+        'canonicalization changed number of triples in graph'
+    assert len(g.all_nodes()) == len(cg.all_nodes()), \
+        'canonicalization changed number of nodes in graph'
+    assert len(list(g.subjects(RDF['type'], RDF['Statement']))) == \
+           len(list(cg.subjects(RDF['type'], RDF['Statement']))), \
+        'canonicalization changed number of statements'
+
+    # counter for subject, predicate and object nodes
+    g_pos_counts = Counter(), Counter(), Counter()
+    for t in g:
+        for i, node in enumerate(t):
+            g_pos_counts[i][t] += 1
+    g_count_signature = [sorted(c.values()) for c in g_pos_counts]
+
+    cg_pos_counts = Counter(), Counter(), Counter()
+    for t in cg:
+        for i, node in enumerate(t):
+            cg_pos_counts[i][t] += 1
+    cg_count_signature = [sorted(c.values()) for c in cg_pos_counts]
+
+    assert g_count_signature == cg_count_signature, \
+        'canonicalization changed node position counts'