We need to be able to identify confluences

[timlinux]

Problem

Given a line layer, identify by selection all confluences (points with more than one upstream nodes and more than zero downstream nodes)

Proposed solution

The implementation should take the following steps:

• extract nodes with basic metadata (upstream count, related line id, related nodes etc.) as per #8
• add an attribute to the node table that indicates if a node is a confluence
• Implement logic that determines confluence on the network based on the logic (in pseudocode):

IF upstream_node_count > 1 AND downstream_node_count > 0
THEN
    confluence = True
ENDIF

The logic here being that if there is more than one upstream node and there are more than zero downstream nodes, so this node must be a meeting point of the rivers

• select all line features in the river layer whose starting node is a confluence
• overlay the nodes layer on the river layer with confluences clearly symbolised (client to provide appropriate symbology, in the mean time we can just make up our own visual vocabulary).

  Expected outcome

• The extracted nodes layer should have a column added called confluence and populated with 0 if the node is not a confluence and 1 if it is.
• The line layer should have all segments selected whose origin node is a confluence selected
• The user should be able to easily visually identify all segments whose origin node is a confluence

Example nodes attribute table (excludes other extracted feature types):

id	upstream_node_count	downstream_node_count	confluence
1	3	0	0
2	2	3	1