Database of geographic place names from OpenStreetMap for full text search downloadable for free. Website: http://osmnames.org
- Downloadable gazeteer data a la GeoNames.org: http://download.geonames.org/export/dump/ but each record has bounding box and an importance rank known from Nominatim
- Data are derived primarily from OpenStreetMap
- The data format is simple to use tab-delimited text in utf8 encoding (as geonames.org). First line has column names.
- Different type of records are stored in different files (download and index just what you need, sometimes you don't need POIs or addresses with house numbers)
- Tight to the OSM2VectorTiles generator (class/type from vector tiles, exporting names available in vector tiles, regular diff updates possible)
- Possible to generate from a country specific extract of Open Street Map (together with vector tiles)
- Direct indexing via fulltext search engines (SphinxSearch, ElasticSearch, etc.)
- Simple process to get a basic search in place names on a map within minutes
- Downloading the basic gazeteer data from OSM in usable format is problematic now (a need to process large OSM Planet files)
- Powered by super fast open-source fulltext Sphinxsearch - inspired by Swiss GeoAdmin search service
- JSON/JSONP API similar to Nominatim: http://nominatim.klokantech.com/?q=paris&format=jsonv2&addressdetails=1
- Ready to use via Docker in minutes
- https://github.com/klokantech/osmnames-sphinxsearch
*osm_id - MUST BE UNIQUE "DOCUMENT ID" accross complete database
display_name - exactly as in Nominatim (may be improved later)
*name (=utf-8)
name_en
name_de
name_es
name_fr
name_ru
name_zh
*class
*type
*north (=boundingbox)
*south
*east
*west
*lat
*lon
scalerank - we have it
place_rank - nominatim has it
importance - exactly as in nominatim calculated
country (=country code, ISO-3166 2-letter country code)
street=<housenumber> <streetname>
city=<city>
county=<county>
state=<state>
country=<country>
(= a la nominatim http://wiki.openstreetmap.org/wiki/Nominatim)
? timestamp - osm modification?
REMARKs:
- Fields like housenumber and postalcode don't belong to this dataset. There's a dataset "OSM Adresses" for that.
You need a complete OSM PBF data dump either from a country extract or of the entire world.
Download the data and put it into the data directory.
wget --directory-prefix=./data http://download.geofabrik.de/europe/switzerland-latest.osm.pbfNow we need to set up the database and import the data using the import-osm Docker container.
# This will automatically initialize the database
docker-compose up -d postgres
# Import the OSM data dump from the ./data folder
docker-compose run import-osmCreate the database schema.
docker-compose run schemaWe can now export the ranked geonames and their geometries.
docker-compose run export-osmnamesThe different components that attach to the postgres container are all located in the src directory.
| Component | Description |
|---|---|
| postgres | PostGIS data store for OSM data and to perform noise analysis |
| import-osm | Imposm3 based import tool with custom mapping to import selective OSM into the database and reconstruct it as GIS geometries |
| export-osmnames | Export names and their bounding boxes to TSV datasets |
| schema | Contains views, tables, functions for the schema |