Spatial Structure

The thinking behind spatial.

In an oracle database, spatial data is stored in a layer, which in simple terms can be considered to be a column in a table where the datatype is SDO_GEOMETRY. This is a simplified view of the model underlying Oracle Spatial.

A layer contains one or more geometries (at least one row in the table).
Geometries consist of one or more elements (at least one entry in the geometry object type).

To take a simple example: A point is represented in the layer as a geometry with one element of type point. Points are usually something indicating a specific position like a point of interest.

More complex shapes and are built up by combining more than one element into a geometry. These may, in the real world, be roads, coastlines, borders, rivers or even continents. These are built up by the combination of many elements, each element being a simple type, giving the geometry. In most cases the elements will follow on from each other to build up a complex geometry, but there is no restriction on a geometry having more than one 'shape', so a geometry could contain the 'shapes' for all the countries in Europe or all the States in the USA.

This means that your GIS application can show maps consisting of points of interest and routes.

To recap: The Spatial Model

Layer --> Geometry --> Element --> (point, line, polygon, compound line, compound polygon).

It is usual practice to keep related geometries in the same layer (column), so the 'roads' (one layer) are not mixed with the 'county borders' (separate layer).