I am editing a road network on one layer, and am looking for a quick way to show the roads that don't intersect (don't meet or overlap). Is SQL the best way to do it? If so I am looking for how to produce the correct code in the 'where Condition' step in the SQL Select Window.
You probably mean "almost intersect", because in a road network there will be a lot of roads which don't intersect at all. The SQL below works on the original + a copy of the table, where you have a unique ID field and a dummy field with 0 for all records. But it gets really slow, even on small tables. The idea is to find objects which don't intersect, but where the buffered objects (5 meter buffer) do intersect.
I would recommend that you turn to software specialized in this, such as RouteFinder from www.routeware.dk, which has been developed by me and some colleagues.
You can do this with SQL.
Now you need to find the records that don't intersect by comparing the ID's to the ID's in the list:
Depending on the amount of data this query might be slow and you might get a better performance by selecting the records that do intersects and then use the Invert Selection tool to select those that don't:
Now click the Invert Selection tool.
This will however only give you a list of roads that don't intersect. There are a number of other possible issues which you really would need a Routing tool to bring forward. RouteFinder (the Pro version) has tools for this built-in.