Function addNewNode() renamed to addNewNode!().

addNewNode!() now starts the search for a new node ID at hash(location) by default, rather than always from 1. This should speed it up and make node IDs more unique.
Note that the hashes are not consistent between julia sessions.
I had to modify the tests slightly, now that the node IDs on boundaries are not guaranteed to have the same IDs between sessions.
See discussion at Issue #65.

Signed-off-by: Ted Steiner <tsteiner2@gmail.com>

src/nodes.jl

@@ -69,8 +69,10 @@ function nodesWithinRange{T<:Union(ENU,ECEF)}(nodes::Dict{Int,T},
 end
 
 ### Add a new node ###
-function addNewNode{T<:Union(LLA,ENU)}(nodes::Dict{Int,T}, loc::T)
-    id = 1
+function addNewNode{T<:Union(LLA,ENU)}(nodes::Dict{Int,T}, 
+                                       loc::T, 
+                                       start_id::Int=abs(int(hash(loc))) )
+    id = start_id
     while id <= typemax(Int)
         if !haskey(nodes, id)
             nodes[id] = loc

test/intersections.jl

@@ -26,7 +26,6 @@ highway_sets = findHighwaySets(hwys)
 # Cluster intersections
 intersection_cluster_mapping = findIntersectionClusters(nodes,intersections,highway_sets,max_dist=15)
 intersection_clusters = unique(collect(values(intersection_cluster_mapping)))
-@test sort!(intersection_clusters) == [1,2,3,4,5]
 @test length(intersection_clusters) == 5
 
 # Replace Nodes in Highways

test/routes.jl

@@ -68,12 +68,14 @@ edges = OpenStreetMap.getEdges(network)
 for k = 1:Graphs.num_edges(network.g)
     @test edges[k].index == k
 end
+#=
 @test edges[10].source.index == 128
 @test edges[10].target.index == 154
 @test edges[20].source.index == 111
 @test edges[30].source.index == 106
 @test edges[40].source.index == 9
 @test edges[50].source.index == 22
+=#
 
 # Form transportation network from segments
 intersections = findIntersections(hwys)
@@ -112,7 +114,7 @@ loc0 = nodesENU[node0]
 filteredENU = filter((k,v)->haskey(network.v,k), nodesENU)
 local_indices = nodesWithinRange(filteredENU, loc0, 100.0)
 @test length(local_indices) == 3
-for index in [61317384, 23, 61317383]
+for index in [61317384, 61317383]
     @test index in local_indices
 end
 
@@ -121,7 +123,7 @@ start_index = nearestNode(filteredENU, loc0)
 node_indices, distances = nodesWithinDrivingDistance(network, start_index, 300.0)
 @test length(node_indices) == length(distances)
 @test length(node_indices) == 14
-for index in [61318574, 18, 17, 575472710, 61317383, 12, 61318436]
+for index in [61318574, 575472710, 61317383, 61318436]
     @test index in node_indices
 end
 for dist in distances
@@ -143,7 +145,7 @@ end
 node_indices, distances = nodesWithinDrivingTime(network, start_index, 50.0)
 @test length(node_indices) == length(distances)
 @test length(node_indices) == 30
-for index in [61318572, 33, 270134895, 575440057, 61323886, 473951349, 986189343]
+for index in [61318572, 270134895, 575440057, 61323886, 473951349, 986189343]
     @test index in node_indices
 end
 for dist in distances
@@ -154,7 +156,7 @@ end
 node_indices, distances = nodesWithinDrivingTime(network, local_indices, 50.0)
 @test length(node_indices) == length(distances)
 @test length(node_indices) == 41
-for index in [575444707, 270134899, 30, 270134936, 986189343]
+for index in [575444707, 270134899, 270134936, 986189343]
     @test index in node_indices
 end
 for dist in distances