Maven Central Build and Test codecov

GeoMesa SQL

Run SQL queries on GeoMesa DataStores, in the command line, or in your programs.

asciicast

Using GeoMesa SQL Programmatically

GeoMesa SQL provides a package for accessing GeoMesa DataStores using JDBC interface. Please refer to the example project for details.

Run SQL using Command Line Tool

GeoMesa SQL comes up with a command line tool geomesa-sqlline, which is a sqlline based utility for running SQL queries on GeoMesa DataStores. It has been tested on the following datastores:

  • Accumulo
  • HBase
  • Cassandra
  • Redis
  • FileSystem

Prerequisites

You need a working GeoMesa command line tool for accessing your GeoMesa DataStore. GeoMesa SQL has been tested with GeoMesa 3.4.1, other versions may also work.

Download the binary distribution

You can download the binary distribution from Releases. Please make sure that the scala version of GeoMesa SQL matches with the GeoMesa command line tool you are using.

Setup GEOMESA_HOME environment variable

GeoMesa SQL command line tool needs jars and config files in GeoMesa command line tool. GEOMESA_HOME must be configured properly before using geomesa-sqlline command. You can configure GEOMESA_HOME in geomesa-sql/conf/geomesa-sql-env.sh, or setting it temporarily in current session by running export GEOMESA_HOME=/path/to/geomesa.

Using GeoMesa SQL command line tool

You can run GeoMesa SQL command line tool using this command:

geomesa-sqlline -c <catalog> [-p <datastore_params>]

For example, to access GeoMesa Accumulo or HBase DataStore in catalog geomesa using default datastore configuration, we can run the following command:

geomesa-sqlline -c geomesa

If you're accessing GeoMesa DataStore requiring more parameters other than catalog, you can specify additional datastore parameters using -p. For example, we need to specify cassandra.contact.point and cassandra.keyspace when accessing GeoMesa Cassandra:

geomesa-sqlline -c geomesa -p "cassandra.contact.point=localhost:9042;cassandra.keyspace=geomesa"

geomesa-sqlline will bring you an interactive shell to run SQL queries. You can list tables in the catalog using !table:

0: jdbc:geomesa:cassandra.catalog=geomesa> !table
+-----------+-------------+------------------------+--------------+---------+----------+------------+
| TABLE_CAT | TABLE_SCHEM |       TABLE_NAME       |  TABLE_TYPE  | REMARKS | TYPE_CAT | TYPE_SCHEM |
+-----------+-------------+------------------------+--------------+---------+----------+------------+
|           | GEOMESA     | beijing_subway         | TABLE        |         |          |            |
|           | GEOMESA     | beijing_subway_station | TABLE        |         |          |            |
|           | GEOMESA     | starbucks              | TABLE        |         |          |            |
|           | metadata    | COLUMNS                | SYSTEM TABLE |         |          |            |
|           | metadata    | TABLES                 | SYSTEM TABLE |         |          |            |
+-----------+-------------+------------------------+--------------+---------+----------+------------+

List columns in a table using !describe "<table_name>":

0: jdbc:geomesa:cassandra.catalog=geomesa> !describe "starbucks"
+-----------+-------------+------------+----------------+-----------+-------------------------------+
| TABLE_CAT | TABLE_SCHEM | TABLE_NAME |  COLUMN_NAME   | DATA_TYPE |           TYPE_NAME           |
+-----------+-------------+------------+----------------+-----------+-------------------------------+
|           | GEOMESA     | starbucks  | __FID__        | 12        | VARCHAR CHARACTER SET "UTF-8" |
|           | GEOMESA     | starbucks  | id_0           | 4         | INTEGER                       |
|           | GEOMESA     | starbucks  | brand          | 12        | VARCHAR CHARACTER SET "UTF-8" |
|           | GEOMESA     | starbucks  | store_name     | 12        | VARCHAR CHARACTER SET "UTF-8" |
|           | GEOMESA     | starbucks  | ownership_type | 12        | VARCHAR CHARACTER SET "UTF-8" |
|           | GEOMESA     | starbucks  | address        | 12        | VARCHAR CHARACTER SET "UTF-8" |
|           | GEOMESA     | starbucks  | city           | 12        | VARCHAR CHARACTER SET "UTF-8" |
|           | GEOMESA     | starbucks  | state_province | 12        | VARCHAR CHARACTER SET "UTF-8" |
|           | GEOMESA     | starbucks  | country        | 12        | VARCHAR CHARACTER SET "UTF-8" |
|           | GEOMESA     | starbucks  | postcode       | 4         | INTEGER                       |
|           | GEOMESA     | starbucks  | geog           | 2015      | GEOMETRY                      |
+-----------+-------------+------------+----------------+-----------+-------------------------------+

Run SQL queries:

0: jdbc:geomesa:cassandra.catalog=geomesa> SELECT store_name, ownership_type, geog FROM starbucks WHERE ownership_type = 'Licensed' LIMIT 10;
+-------------------------------------+----------------+-----------------------------+
|             STORE_NAME              | OWNERSHIP_TYPE |            GEOG             |
+-------------------------------------+----------------+-----------------------------+
| Safeway-Ewa #2897                   | Licensed       | POINT (-158.01933 21.33606) |
| Hilton Hawaiian Village-Kalia Tower | Licensed       | POINT (-157.83602 21.28403) |
| Target Kailua T-2697                | Licensed       | POINT (-157.7407 21.39112)  |
| Schofield Barracks Main Store Mall  | Licensed       | POINT (-158.07508 21.47854) |
| Safeway-Lihue #2894                 | Licensed       | POINT (-159.38637 21.96841) |
| Safeway - Kihei #1500               | Licensed       | POINT (-156.449 20.75245)   |
| Marriott's Maui Ocean Club          | Licensed       | POINT (-156.69526 20.9167)  |
| Ralphs-Palos Verdes #720            | Licensed       | POINT (-118.40603 33.74952) |
| Sodexo@Naval Station 32nd. St. San  | Licensed       | POINT (-117.12923 32.69084) |
| Vons-Chula Vista #2071              | Licensed       | POINT (-116.9393 32.64391)  |
+-------------------------------------+----------------+-----------------------------+
10 rows selected (0.066 seconds)

Run SQL queries with spatial functions:

0: jdbc:geomesa:cassandra.catalog=geomesa> SELECT store_name, geog FROM starbucks WHERE ST_Within(geog, ST_GeomFromText('POLYGON((-71.78681142189922 42.84501461908715,-70.90446657994212 41.33088304889827,-69.63607675888956 42.76409278843215,-71.78681142189922 42.84501461908715))')) LIMIT 10;
+------------------------------+----------------------------+
|          STORE_NAME          |            GEOG            |
+------------------------------+----------------------------+
| Taunton Depot                | POINT (-71.0667 41.87916)  |
| Target Easton T-2267         | POINT (-71.14357 42.02736) |
| Brockton, Westgate Drive     | POINT (-71.0543 42.09391)  |
| Stop & Shop-Plymouth #469    | POINT (-70.68628 41.95501) |
| Hingham, Rte. 53             | POINT (-70.9013 42.18106)  |
| Target Taunton T-1189        | POINT (-71.06911 41.8783)  |
| Target Plainville T-1930     | POINT (-71.31021 42.03397) |
| Medfield, Main St & North St | POINT (-71.30541 42.18706) |
| Target Wareham T-2292        | POINT (-70.74462 41.7746)  |
| Brockton, Belmont Street     | POINT (-71.06865 42.05886) |
+------------------------------+----------------------------+
10 rows selected (0.095 seconds)

If you want to query tables scattered in multiple catalogs, you can write a Calcite model file and run geomesa-sql -m /path/to/calcite-model-file.yaml. Here is an example of Calcite model file:

version: 1.0
defaultSchema: catalog1
schemas:
  - name: catalog1
    type: custom
    factory: com.spatialx.geomesa.sql.GeoMesaSchemaFactory
    operand:
      # Options for connecting to GeoMesa DataStore
      "hbase.catalog": "geomesa_catalog1"

  - name: catalog2
    type: custom
    factory: com.spatialx.geomesa.sql.GeoMesaSchemaFactory
    operand:
      "hbase.catalog": "geomesa_catalog2"

Building from Source

Requirements:

Run this command to download the source repository and build this project:

git clone [email protected]:bxkftechteam/geomesa-sql.git
cd geomesa-sql
mvn clean package

This will build the artifacts and run unit tests using a mini Accumulo cluster. You can run mvn clean package -DskipTests to skip running tests. The distribution package containing the geomesa-sqlline tool will be generated in geomesa-sql-dist/target directory.

This project depends on Scala 2.12 by default. You can run ./build/change-scala-version.sh 2.11 to switch dependency to Scala 2.11.

After building the project from source, you can test and debug geomesa-sqlline directly in geomesa-sql-cli directory.

cd geomesa-sql-cli
GEOMESA_HOME=/path/to/geomesa ./bin/geomesa-sqlline -c <catalog>

Contributing

This project is still in its early stage, issues and pull requests are welcomed. If you are interested in contributing to this project please read the Contribution Guide.