How to use EPSG geodetic dataset

The EPSG geodetic dataset is a de-facto standard providing thousands of Coordinate Reference System (CRS) definitions together with information about how to perform coordinate operations, their accuracies and their domains of validity. The EPSG dataset is owned and maintained by the International Association of Oil & Gas producers. Usage of EPSG dataset with Apache SIS is optional but strongly recommended: without that geodetic dataset, only a small subset of CRS definitions will be available (basically the constants enumerated in the CommonCRS Java class) unless full definitions are provided in Well Known Text (WKT) or Geographic Markup Language (GML) formats. Furthermore, coordinate operations between any given pair of CRS may be less accurate and their domains of validity may be unspecified if Apache SIS cannot query EPSG.

The EPSG geodetic dataset is not distributed with Apache SIS because the EPSG terms of use are incompatible with Apache license. The following items are quoted from those terms of use:

  • The EPSG Facilities are published by IOGP at no charge. Distribution for profit is forbidden.
  • The data may be included in any commercial package provided that any commerciality is based on value added by the provider and not on a value ascribed to the EPSG Dataset which is made available at no charge.
  • Ownership of the EPSG Dataset by IOGP must be acknowledged in any publication or transmission (by whatever means) thereof (including permitted modifications).
  • Modification of parameter values is permitted as described in the table 1 to allow change to the content of the information provided that numeric equivalence is achieved.
  • No data that has been modified other than as permitted in these Terms of Use shall be attributed to the EPSG Dataset.

In order to use the EPSG geodetic dataset with Apache SIS, apply one of the following choices:

Install a local copy with command-line tool

The installation process described in this section makes clear that EPSG dataset is distributed under a different license and asks users for their agreement. This process can be used with the Apache SIS 1.3 binary bundle. If the command-line tool has been downloaded and installed, just query any CRS. For example:

sis crs EPSG:6676

Alternatively if the JavaFX application is used, just start the application. It may be necessary to open a random data file for triggering the EPSG dataset initialization. In both cases, the first time that the command-line tool or JavaFX application needs to query EPSG, it will prompt the user for authorization to download EPSG geodetic dataset from Maven Central. If the user accepts EPSG terms of use, then a local copy of the EPSG geodetic dataset will be created and stored in the apache-sis-1.3/data sub-directory.

How to use the local copy in other applications

The EPSG dataset installed by the command-line tools or the JavaFX application can also be used with other applications. For using the installed EPSG geodetic dataset in your own application, apply one of the following choices:

  • Set the SIS_DATA environment variable to the path of apache-sis-1.3/data directory (recommended).
  • Set the derby.system.home Java property to the path of apache-sis-1.3/data/Databases directory.

Examples are shown below for Unix systems, assuming that the current directory is the directory where apache-sis-1.3-bin.zip has been unzipped (replace myApp and MyMainClass by the application to launch):

export SIS_DATA=apache-sis-1.3/data
java --class-path apache-sis-1.3/lib/sis-referencing.jar:myApp.jar MyMainClass

If the SIS_DATA environment variable cannot be set, Java property can be used as a fallback:

java -Dderby.system.home=apache-sis-1.3/data/Databases \
     --class-path apache-sis-1.3/lib/sis-referencing.jar:myApp.jar \
     MyMainClass

Alternatively SIS_DATA or derby.system.home can be set to the path of any other directory which contain the same files.

Add a Maven dependency

Maven projects can get the EPSG geodetic dataset automatically, without any prompt for terms of use agreement, if they add a sis-epsg or sis-embedded-data dependency (from org.apache.sis.non-free group) in their project. Those two approaches have advantages and inconvenient described in following sub-sections. In both cases, we assume that developers who add those dependencies explicitly in their project agree with EPSG terms of use.

As database installer

With sis-epsg artifact on the classpath, Apache SIS will create a local copy of EPSG database when first needed. The target database must be specified by users with one of the following choices:

  • Set the SIS_DATA environment variable to the path of an initially empty directory (recommended). The specified directory must exist, but sub-directories will be created as needed.
  • Set the derby.system.home Java property to the path of an initially empty directory, or a directory that contain other Derby databases. The specified directory must exist.
  • Register a DataSource under the java:comp/env/jdbc/SpatialMetadata name in a JNDI directory (see next section). The database must exist but can be initially empty.
  • Set a DataSource from Java code.

The Maven dependency is as below (the Derby dependency can be replaced by another database driver if that database is specified by JNDI):

<dependencies>
  <dependency>
    <groupId>org.apache.sis.non-free</groupId>
    <artifactId>sis-epsg</artifactId>
    <version>1.3</version>
    <scope>runtime</scope>
  </dependency>
  <dependency>
    <groupId>org.apache.derby</groupId>
    <artifactId>derby</artifactId>
    <version>10.14.2.0</version>
    <scope>runtime</scope>
  </dependency>
</dependencies>

See the download page for more information about Maven dependency declaration.

As embedded database

With sis-embedded-data artifact on the classpath, there is no need to setup environment variable, Java property or JNDI. However this simplicity come with the following inconvenient:

  • a larger download,
  • no option for choosing which data to use (and consequently which license to accept),
  • no possibility to choose the database engine (i.e. the database software is fixed to Derby),
  • no possibility to add user data (i.e. the database is read-only),
  • slower execution of CRS.forCode(…) and CRS.findCoordinateOperation(…) methods, unless the JAR file is uncompressed.

This dependency can be declared as below (see the download page for more information about Maven dependency declaration). Note that sis-epsg and sis-embedded-data should not be specified in the same project; only one is needed:

<dependencies>
  <dependency>
    <groupId>org.apache.sis.non-free</groupId>
    <artifactId>sis-embedded-data</artifactId>
    <version>1.3</version>
    <scope>runtime</scope>
  </dependency>
</dependencies>

The performance issue can be avoided if the JAR file is uncompressed. But uncompressed sis-embedded-data.jar file is more than 5 times larger than the compressed file. Given that CRS​.forCode(…) and CRS​.findCoordinateOperation(…) should not be invoked too often, and that performance degradation does not apply to the CoordinateOperation instances created by those method calls, the JAR file is distributed on the Maven repository in its compressed form. If desired, better performance can be achieved by using one of the other configurations described in this page, or by uncompressing the sis-embedded-data.jar file locally.

Use an existing EPSG database

Applications can use their own EPSG database. In addition of Derby, Apache SIS is also tested on HSQL and PostgreSQL. For using an arbitrary database, register a javax.sql.DataSource instance through one of the methods described below. The database must exist but can be empty, in which case it will be populated with an EPSG schema when first needed if the org.apache.sis.non-free:​sis-epsg:​1.3 dependency is on the classpath (see above section). If non-empty, the database should contain the tables created by SQL scripts downloaded from EPSG. Note thas as of Apache SIS 1.2, only EPSG dataset version 9 is supported. EPSG datasets version 10 and later are not yet supported.

Registration by Java code

The data source can be specified by Java code as below (replace the main method by any method where initialization occurs):

import javax.sql.DataSource;
import org.apache.sis.setup.Configuration;

public void MyApp {
    public static void main(String[] args) {
        Configuration.current().setDatabase(MyApp::createDataSource);
    }

    private static DataSource createDataSource() {
        DataSource ds = ...;        // Initialize the data source here.
        return ds;
    }
}

Registration by Java Naming and Directory Interface

Registration in JNDI can be done programmatically (by Java code) or by configuring XML files in some environments. Registration can be done by the following Java code, provided that a JNDI implementation is available on the classpath:

// Example using PostgreSQL data source (org.postgresql.ds.PGSimpleDataSource)
PGSimpleDataSource ds = new PGSimpleDataSource();
ds.setDatabaseName("SpatialMetadata");
// Server default to "localhost".

// Registration assuming that a JNDI implementation is available
Context env = (Context) InitialContext.doLookup("java:comp/env");
env.bind("jdbc/SpatialMetadata", ds);

If there is no JNDI environment, the org.apache.sis.setup.Configuration class can be used as a fallback:

// Fallback if no JNDI environment is available.
Configuration.current().setDatabase(() -> ds);

Registration in web application containers

JNDI implementations are provided by web application containers like Apache Tomcat. When Apache SIS is used in a JavaEE container, the data source can be configured as below:

  1. Make the JDBC driver available to the web container and its applications. On Tomcat, this is accomplished by installing the driver’s JAR files into the $CATALINA_HOME/lib directory.

  2. If using Derby, copy derby.war into the $CATALINA_HOME/webapps directory and specify the directory where the Derby databases are located (skip this step if another database is used):

export JAVA_OPTS=-Dderby.system.home=$SIS_DATA/Databases
  1. Declare the JNDI name in application WEB-INF/web.xml file:
<resource-ref>
  <description>EPSG dataset and other metadata used by Apache SIS.</description>
  <res-ref-name>jdbc/SpatialMetadata</res-ref-name>
  <res-type>javax.sql.DataSource</res-type>
  <res-auth>Container</res-auth>
</resource-ref>
  1. Configure the data source in $CATALINA_HOME/conf/context.xml or in application META-INF/context.xml file (change attribute values as needed for the chosen JDBC driver):
<Context crossContext="true">
  <WatchedResource>WEB-INF/web.xml</WatchedResource>
  <Resource name            = "jdbc/SpatialMetadata"
            auth            = "Container"
            type            = "javax.sql.DataSource"
            username        = "sa"
            password        = "sa"
            driverClassName = "org.apache.derby.jdbc.EmbeddedDriver"
            url             = "jdbc:derby:SpatialMetadata"/>
</Context>
  1. If using Derby, verify on the localhost:8080/derby/derbynet page (skip this step if another database is used).

More advanced configurations are possible. For example Tomcat can invoke a custom Java method instead of fetching the data source from the context.xml file.