geofdw is a collection of PostGIS-related
foreign data
wrappers for
PostgreSQL written in Python using the
multicorn extension. By using a FDW, you can
access spatial data through Postgres tables without having to import the
data first, which can be useful for dynamic or non-tabular data
available through web services.
Currently implemented forward data wrappers are:
- FGeocode: forward geocoding
- RGeocode: reverse geocoding
- GeoJSON: online GeoJSON
- RandomPoint: random point in a bounding box
geofdw uses plpygis to
manipulate PostGIS geometries.
Create a single server for all forward geocoding tables:
# CREATE SERVER fwd_geocode FOREIGN DATA WRAPPER multicorn OPTIONS ( wrapper 'geofdw.FGeocode' );
Create two tables, one using the GoogleV3 geocoder and one using Nominatim:
# CREATE FOREIGN TABLE fgc_google ( query TEXT, rank INTEGER, address TEXT, geom GEOMETRY ) SERVER fwd_geocode; # CREATE FOREIGN TABLE fgc_nominatim ( query TEXT, rank INTEGER, address TEXT, geom GEOMETRY ) SERVER fwd_geocode OPTIONS ( service 'nominatim');
Select results from the geocoder matching our query string:
SELECT address, ST_AsText(geom) AS geom FROM fgc_google WHERE query = 'canada house' LIMIT 5;
address | geom
----------------------------------------------------------------------------------------+------------------------------------
Canada House, London, UK | POINT Z (-0.1291 51.5077 0)
Canada House, Temple Road, Blackrock, Co. Dublin, Ireland | POINT Z (-6.1756563 53.2994401 0)
Canada House, Saint Stephen's Green, Dublin 2, Ireland | POINT Z (-6.2576992 53.335963 0)
Canada House, 29 Hampton Road, Twickenham, Greater London TW2 5QE, UK | POINT Z (-0.3443802 51.441739 0)
Canada House, 272 Field End Road, Ruislip, Greater London HA4 9NA, UK | POINT Z (-0.3973435 51.5752994 0)
Perform the same query but using the Nominatim geocoder:
# SELECT address, ST_AsText(geom) AS geom FROM fgc_nominatim WHERE query = 'canada house' LIMIT 5;
address | geom
-----------------------------------------------------------------------------------------------------------------------------------------------+--------------------------------------------
High Commission of Canada, 5, Trafalgar Square, Covent Garden, City of Westminster, London, Greater London, England, SW1Y 5BJ, United Kingdom | POINT Z (-0.129104703393283 51.50782475 0)
Canada House, West 54th Street, Diamond District, Manhattan, New York, NYC, New York, 10019, United States of America | POINT Z (-73.9758789212845 40.7609797 0)
Canada House, Kololo Road, Governments Cantonment, Juba, Central Equatoria, South Sudan | POINT Z (31.5901769 4.8615639 0)
Canada House, Justine Close, Nalumunye, Kazinga, Wakiso, Central 2, Central Region, Uganda | POINT Z (32.4868336484328 0.25676655 0)
בית קנדה, 1, שבי ציון, רובע א', אשדוד, מחוז הדרום, 77452, מדינת ישראל | POINT Z (34.64463585 31.8086878 0)
Influence the geocoder by restricting the query to a certain bounding box (in this case, without the hint, Google will only return results in the USA):
# SELECT address, ST_AsText(geom) AS geom FROM fgc_google WHERE query = 'Water St' AND geom && ST_GeomFromEWKT('SRID=4326;POLYGON((50 2, 55 2, 55 -2, 50 -2, 50 2))');
address | geom
------------------------------------------------------------------------+-----------------------------------
Water Street, Lavenham, Sudbury, Suffolk CO10 9RW, UK | POINT Z (0.7982752 52.1071416 0)
Water Street, Stamford, Lincolnshire PE9 2NJ, UK | POINT Z (-0.4752917 52.649958 0)
Water Street, Cambridge, Cambridgeshire CB4 1PA, UK | POINT Z (0.1474321 52.2184911 0)
Water Street, Hampstead Norreys, Thatcham, West Berkshire RG18 0RU, UK | POINT Z (-1.2408465 51.4854726 0)
Water Street, Burntwood, Staffordshire WS7 1AW, UK | POINT Z (-1.9355616 52.6839693 0)
Water Street, Kettering, Northamptonshire NN16, UK | POINT Z (-0.7173979 52.4008413 0)
Water Street, Birmingham, West Midlands B3 1HP, UK | POINT Z (-1.9028035 52.4854385 0)
Water Street, London WC2R 3LA, UK | POINT Z (-0.1136366 51.5118691 0)