Unofficial Go port of the Geo Library for Amazon DynamoDB (https://github.com/amazon-archives/dynamodb-geo) using geohash (https://en.wikipedia.org/wiki/Geohash) to easily create and query geospatial data.
Unofficial Go port of the Geo Library for Amazon DynamoDB using geohash to easily create and query geospatial data.
The library takes care of managing the geohash indexes and storing item with latitude/longitude pairs.
Fetch the package with
go get github.com/crolly/dyngeo
And import it into your programs with
import "github.com/crolly/dyngeo"
type DynGeoConfig struct {TableName stringConsistentRead boolHashKeyAttributeName stringRangeKeyAttributeName stringGeoHashAttributeName stringGeoJSONAttributeName stringGeoHashIndexName stringHashKeyLength int8LongitudeFirst boolDynamoDBClient *dynamodb.DynamoDB}
Defines, how DynG(e)o manages the geospatial data, e.g. what the db attribute and index names are as well as setting the geohash key length.
The geohash key length will determine the size of the tiles the planet will be seperated into:
| Length | Tile Size |
|---|---|
| 1 | 5,009.4km x 4,992.6km |
| 2 | 1,252.3km x 624.1km |
| 3 | 156.5km x 156km |
| 4 | 39.1km x 19.5km |
| 5 | 4.9km x 4.9km |
| 6 | 1.2km x 609.4m |
| 7 | 152.9m x 152.4m |
| 8 | 38.2m x 19m |
| 9 | 4.8m x 4.8m |
| 10 | 1.2m x 59.5cm |
| 11 | 14.9cm x 14.9cm |
| 12 | 3.7cm x 1.9cm |
Setting DynamoDBClient *dynamodb.DynamoDB and TableName string is required.
func New(config DynGeoConfig) (*DynGeo, error)
Returns a new instance of DynG(e)o managing the geohashing and geospatial db operations.
func (dg DynGeo) PutPoint(input PutPointInput) (*PutPointOutput, error)
Put a point into the Amazon DynamoDB table. Once put, you cannot update attributes specified in GeoDataManagerConfiguration: hash key, range key, geohash and geoJson. If you want to update these columns, you need to insert a new record and delete the old record.
func (dg DynGeo) BatchWritePoints(inputs []PutPointInput) (*BatchWritePointOutput, error)
Put a list of points into the Amazon DynamoDB table. Once put, you cannot update attributes specified in GeoDataManagerConfiguration: hash key, range key, geohash and geoJson. If you want to update these columns, you need to insert a new record and delete the old record.
func (dg DynGeo) GetPoint(input GetPointInput) (*GetPointOutput, error)
Get a point from the Amazon DynamoDB table.
func (dg DynGeo) UpdatePoint(input UpdatePointInput) (*UpdatePointOutput, error)
Update a point data in Amazon DynamoDB table. You cannot update attributes specified in GeoDataManagerConfiguration: hash key, range key, geohash and geoJson. If you want to update these columns, you need to insert a new record and delete the old record.
func (dg DynGeo) DeletePoint(input DeletePointInput) (*DeletePointOutput, error)
Delete a point from the Amazon DynamoDB table.
func (dg DynGeo) QueryRadius(input QueryRadiusInput, out interface{}) error
Query a circular area constructed by a center point and its radius.
func (dg DynGeo) QueryRectangle(input QueryRectangleInput, out interface{}) error
Query a rectangular area constructed by two points and return all points within the area. Two points need to construct a rectangle from minimum and maximum latitudes and longitudes. If minPoint.Longitude > maxPoint.Longitude, the rectangle spans the 180 degree longitude line.
This repository contains a Getting Started example in the folder starbucks-example inspired by James Beswick’s very good blog post about Location-based search results with DynamoDB and Geohash
It uses the US Starbucks locations, loads them into DynamoDB in batches of 25 and then retrieves the the locations of all Starbucks in the radius of 5000 meters surrounding Latitude: 40.7769099, Longitude: -73.9822532.
The example illustrates the general usage as well as a hint of the performance. The radius search constantly needs about 20-30ms with the given dataset (approximately 6500 Starbucks coffee shops in the US).