* Ubuntu: `sudo apt-get install python-gdal unzip`
* CentOS: `sudo yum install gdal-python unzip`
- 2. Get preprocessed TIGER 2015 data and unpack it into the
+ 2. Get preprocessed TIGER 2017 data and unpack it into the
data directory in your Nominatim sources:
cd Nominatim/data
SQL files, Nominatim needs for the import:
1. Get the TIGER 2017 data. You will need the EDGES files
- (3,234 zip files, 11GB total). Choose one of the two sources:
+ (3,234 zip files, 11GB total).
wget -r ftp://ftp2.census.gov/geo/tiger/TIGER2017/EDGES/
- wget -r ftp://mirror1.shellbot.com/census/geo/tiger/TIGER2017/EDGES/
-
- The first one is the original source, the second a considerably faster
- mirror.
2. Convert the data into SQL statements:
SELECT place_id, name_vector, address_rank, search_rank,
ST_Distance(centroid, point) as distance, null as isguess
FROM search_name_-partition-
- WHERE name_vector @> isin_token
+ WHERE name_vector && isin_token
AND ST_DWithin(centroid, point, 0.015)
AND search_rank between 26 and 27
ORDER BY distance ASC limit 1
SELECT place_id, name_vector, address_rank, search_rank,
ST_Distance(centroid, point) as distance, null as isguess
FROM search_name_-partition-
- WHERE name_vector @> isin_token
+ WHERE name_vector && isin_token
AND ST_DWithin(centroid, point, 0.03)
AND search_rank between 16 and 22
ORDER BY distance ASC limit 1
| object | parent_place_id |
| N1 | W2 |
+ # github #1056
+ Scenario: Full names should be preferably matched for nearest road
+ Given the grid
+ | 1 | | 2 | 5 |
+ | | | | |
+ | 3 | | | 4 |
+ | | 10| | |
+ And the places
+ | osm | class | type | name+name | geometry |
+ | W1 | highway | residential | Via Cavassico superiore | 1, 2 |
+ | W3 | highway | residential | Via Cavassico superiore | 2, 5 |
+ | W2 | highway | primary | Via Frazione Cavassico | 3, 4 |
+ And the named places
+ | osm | class | type | addr+street |
+ | N10 | shop | yes | Via Cavassico superiore |
+ When importing
+ Then placex contains
+ | object | parent_place_id |
+ | N10 | W1 |