|Version 16 (modified by 10 years ago) ( diff ),|
For portability, the default storage for Features is the normal database via Web2Py's DAL.
In order to support Lines & Polygons we store as WKT
- UI for Points can hide this behind simple Lat/Lon
- Shapely used to do conversion
We can extend DAL for spatial support:
- adding if deployment_settings.gis.spatialdb == True and deployment_settings.database.db_type == "postgres" to routines inside S3GIS' spaital functions to use optimised routines where possible & falling back to the existing Shapely routines.
response.custom_commit = lambda: do_commit()
Since we store features in Lat/Lon then we need to use the geography_columns, which stay updated automatically :)
geog GEOGRAPHY(Point) geog GEOGRAPHY(Polygon) ST_Distance(geography, geography) returns double ST_DWithin(geography, geography, float8) returns boolean ST_Area(geography) returns double ST_Length(geography) returns double ST_Covers(geography, geography) returns boolean ST_CoveredBy(geography, geography) returns boolean ST_Intersects(geography, geography) returns boolean ST_Centroid(geometry) returns a point that is approximately on the center of mass of the input argument. This simple calculation is very fast, but sometimes not desirable, because the returned point is not necessarily in the feature itself. If the input feature has a convexity (imagine the letter ‘C’) the returned centroid might not be in the interior of the feature. ST_PointOnSurface(geometry) returns a point that is guaranteed to be inside the input argument. It is substantially more computationally expensive than the centroid operation
Q: integrate FeatureServer?
Q: copy GeoDjango?
Extend DAL to support Spatial Queries
We need to support some extra SQL queries in DAL.
I guess we have a PostGIS Adapter which inherits from the PostgreSQL one & a Spatialite adapter which inherits from the Sqlite one.
Both should inherit the OpenGIS set of SQL syntax:
We need to be able to do spatial queries like this:
This one is our key performance bottleneck currently as we do BBOX filters on GeoJSON feature layers like this:
Instead of this query:
bbox_filter = ((table.lon > minLon) & (table.lon < maxLon) & (table.lat > minLat) & (table.lat < maxLat))
We want to:
bbox_filter = (table.ST_Overlaps([minLon, minLat, maxLon, maxLat]))
Full list of possible commands are here, but I don't expect all these to be implemented immediately (we can add additional ones as/when we need them if the hooks are right):
We also want to be able to specify that a table is spatialised, so an option to db.define_table(spatial=True). If this is on then we should send AddGeometryColumn() to the SQL Adapter
NB A normal system would commonly only have a single spatial table, so this could be easily done out of band using a script, as I do now:
There are also some new field types, but I don't think we need to worry about those right now.