Add graph hashing method to utils

buildingmotif/utils.py

@@ -513,3 +513,23 @@ def skip_uri(uri: URIRef) -> bool:
         if uri.startswith(ns):
             return True
     return False
+
+
+def hash_graph(graph: Graph) -> int:
+    """
+    Returns a cryptographic hash of the graph contents
+
+    :param graph: graph to hash
+    :type graph: graph
+
+    :return: integer hash
+    :rtype: int
+    """
+    # Copy graph to memory (improved performance if graph is backed by a DB store)
+    graph_prime = copy_graph(graph)
+
+    # nt is the best performing serialization format I tested.
+    # For medium-office-compiled it takes 0.03s vs 0.5 for ttl
+    graph_string = graph_prime.serialize(format="nt")
+
+    return hash(graph_string)

tests/unit/test_utils.py

@@ -10,6 +10,7 @@
     _param_name,
     get_parameters,
     get_template_parts_from_shape,
+    hash_graph,
     replace_nodes,
     rewrite_shape_graph,
     skip_uri,
@@ -278,3 +279,22 @@ def test_skip_uri():
     assert skip_uri(XSD.integer)
     assert skip_uri(SH.NodeShape)
     assert not skip_uri(BRICK.Sensor)
+
+
+def test_hash():
+    graph = Graph()
+    graph.parse(data=PREAMBLE)
+
+    graph.add((MODEL["a"], A, BRICK["AHU"]))
+    before_hash = hash_graph(graph)
+
+    triple_to_add = (MODEL["b"], A, BRICK["Sensor"])
+    graph.add(triple_to_add)
+
+    after_hash = hash_graph(graph)
+    assert before_hash != after_hash, "Graph changed, but hashes did not"
+
+    graph.remove(triple_to_add)
+
+    after_hash = hash_graph(graph)
+    assert before_hash == after_hash, "Graph with same state resulted in different hash"