Improved has_overlaps checks and tests

tests/test_functions.py

@@ -414,11 +414,11 @@ def test_get_path_length():
 
 def test_has_overlaps():
     edges = [
-        Edge(0, 1, "A", {"EDGE_ID": 1}),
-        Edge(1, 2, "B", {"EDGE_ID": 2}),
-        Edge(2, 3, "C", {"EDGE_ID": 3}),
-        Edge(2, 3, "C", {"EDGE_ID": 2}),
-        Edge(2, 3, "C", {"EDGE_ID": 4}),
+        Edge(0, 1, "A", {"EDGE_ID": 1, "OFFSET": 0, "LEN_": 10}),
+        Edge(1, 2, "B", {"EDGE_ID": 2, "OFFSET": 0, "LEN_": 10}),
+        Edge(2, 3, "C", {"EDGE_ID": 3, "OFFSET": 0, "LEN_": 10}),
+        Edge(2, 3, "C", {"EDGE_ID": 2, "OFFSET": 0, "LEN_": 10}),
+        Edge(2, 3, "C", {"EDGE_ID": 4, "OFFSET": 0, "LEN_": 10}),
     ]
     assert functions.has_overlaps(edges) is True
 
@@ -444,6 +444,17 @@ def test_has_no_overlaps_loop():
     assert functions.has_overlaps(edges) is False
 
 
+def test_has_no_overlaps_loop_with_split():
+    edges = [
+        Edge(0, 1, "A", {"EDGE_ID": 1, "OFFSET": 5, "LEN_": 10}),
+        Edge(1, 2, "B", {"EDGE_ID": 2, "OFFSET": 0, "LEN_": 10}),
+        Edge(1, 2, "B", {"EDGE_ID": 2, "OFFSET": 10, "LEN_": 10}),
+        Edge(2, 3, "C", {"EDGE_ID": 3, "OFFSET": 0, "LEN_": 10}),
+        Edge(3, 0, "C", {"EDGE_ID": 4, "OFFSET": 0, "LEN_": 5}),
+    ]
+    assert functions.has_overlaps(edges) is False
+
+
 def test_doctest():
     import doctest
 
@@ -456,9 +467,9 @@ def test_doctest():
     # test_edges_to_graph()
     # test_get_edges_from_nodes()
     # test_edges_to_graph()
-    test_get_shortest_edge()
-    test_get_shortest_edge_identical()
-    test_get_shortest_edge_mixed_keys()
+    # test_get_shortest_edge()
+    # test_get_shortest_edge_identical()
+    # test_get_shortest_edge_mixed_keys()
     # test_get_unique_ordered_list()
     # test_get_edges_from_node_pair()
     # test_to_edge()
@@ -471,4 +482,6 @@ def test_doctest():
     # test_get_edges_from_nodes_non_unique()
     # test_has_no_overlaps()
     # test_has_no_overlaps_loop()
+    # test_has_overlaps()
+    test_has_no_overlaps_loop_with_split()
     print("Done!")

wayfarer.pyproj

@@ -4,7 +4,7 @@
     <SchemaVersion>2.0</SchemaVersion>
     <ProjectGuid>f8ac0630-98af-462b-b7ee-5688d6183c85</ProjectGuid>
     <ProjectHome>.</ProjectHome>
-    <StartupFile>tests\test_splitting.py</StartupFile>
+    <StartupFile>tests\test_functions.py</StartupFile>
     <SearchPath>
     </SearchPath>
     <WorkingDirectory>.</WorkingDirectory>
@@ -144,6 +144,7 @@
     <Content Include="dev_setup.ps1" />
     <Content Include="build_docs.ps1" />
     <Content Include="LICENSE" />
+    <Content Include="requirements.demo.txt" />
     <Content Include="run_local.ps1" />
     <Content Include="README.rst" />
     <Content Include="requirements.txt">

wayfarer/__init__.py

@@ -1,7 +1,7 @@
 from typing import NamedTuple
 
 
-__version__ = "0.9.3"
+__version__ = "0.9.4"
 
 LENGTH_FIELD = "LEN_"
 EDGE_ID_FIELD = "EDGE_ID"

wayfarer/functions.py

@@ -16,6 +16,7 @@
     NODEID_FROM_FIELD,
     NODEID_TO_FIELD,
     WITH_DIRECTION_FIELD,
+    OFFSET_FIELD,
     Edge,
 )
 
@@ -612,27 +613,33 @@ def get_multiconnected_nodes(edges, connections=2):
 def has_overlaps(edges: list[Edge]) -> bool:
     """
     Checks if the provided list of edges has overlaps
-    where the same edge id is repeated in the sequence.
-    Edge ids can be adjacent e.g. 1,2,2,3 returns False
-    and the same edge id can be found at the start and end of the sequence
-    e.g. 1,2,3,1 returns False
+    If an edge is split but the measures don't overlap then this will return False
+    The same edge id can be found at the start and end of the sequence if there is
+    no overlap of measures
 
-    >>> edges = [Edge(0, 1, "A", {"EDGE_ID": 1}), Edge(1, 2, "B", {"EDGE_ID": 2})]
+    >>> edges = [Edge(0, 1, "A", {"EDGE_ID": 1, "OFFSET": 0, "LEN_": 10}), \
+Edge(1, 2, "B", {"EDGE_ID": 2, "OFFSET": 0, "LEN_": 10})]
     >>> has_overlaps(edges)
     False
     """
     # get the edge id from the attributes rather than the edge key in case
     # there are split edges which will have modified keys
 
-    edge_ids = [e.attributes[EDGE_ID_FIELD] for e in edges]
+    sorted_edges = sorted(edges, key=lambda e: e.attributes[EDGE_ID_FIELD])
+    grouped_edges = itertools.groupby(
+        sorted_edges, key=lambda e: e.attributes[EDGE_ID_FIELD]
+    )
 
-    grouped_edge_ids = [k for k, g in itertools.groupby(edge_ids)]
-    unique_edge_count = len(set(edge_ids))
+    for _, edge_group in grouped_edges:
+        for edge1, edge2 in itertools.combinations(edge_group, 2):
 
-    if edge_ids[0] == edge_ids[-1]:
-        unique_edge_count += 1  # allow looping back onto start edge
+            from_m1 = edge1.attributes.get(OFFSET_FIELD, 0)
+            to_m1 = from_m1 + edge1.attributes.get(LENGTH_FIELD, 0)
 
-    if len(grouped_edge_ids) > unique_edge_count:
-        return True
-    else:
-        return False
+            from_m2 = edge2.attributes.get(OFFSET_FIELD, 0)
+            to_m2 = from_m2 + edge2.attributes.get(LENGTH_FIELD, 0)
+
+            if max(from_m1, from_m2) < min(to_m1, to_m2):
+                return True
+
+    return False