Skip to content

Latest commit

 

History

History
97 lines (60 loc) · 2.5 KB

DEVELOPING.md

File metadata and controls

97 lines (60 loc) · 2.5 KB
Raw

JTS Developing Guide

Building

The JTS build chain uses Maven. Build commands are executed at the project root directory (this one).

  • Build JTS, with tests:

      mvn clean install

  • Build JTS, no tests:

      mvn clean install -DskipTests

  • Build jts-io-ora:

      mvn install -Poracle

  • Build jts-io-sde:

      mvn install -Parcsde

  • Build everything:

      mvn install -Pall

Javadoc

  • Build Javadoc for core modules

      mvn javadoc:aggregate

Eclipse Configuration

  • Generate Eclipse configuration using mvn eclipse:eclipse
  • Import the generated projects into an Eclipse workspace

Run Configurations

It is convenient to define the following Run Configurations:

  • JTS TestRunner - for executing XML tests:
Field Value
Type Java Application
Project jts-tests
Main class org.locationtech.jtstest.testrunner.JTSTestRunnerCmd
Program arguments -files validate general
Working directory ${workspace_loc:jts-tests/src/test/resources/testxml}
  • JTS TestBuilder - for viewing and processing geometry with JTS
Field Value
Type Java Application
Project jts-app
Main class org.locationtech.jtstest.testbuilder.JTSTestBuilder
Program arguments (optional) -geomfunc <classname> ...
VM args -Xmx1000M
VM args (optional, for Mac) -Dswing.defaultlaf=javax.swing.plaf.metal.MetalLookAndFeel
Working directory Default

Testing

JTS aims for 100% code coverage for unit tests.

There are two kinds of unit tests:

JUnit tests

Used for verifying API code, internal data structures, and ancillary algorithms. These tests are written in Java.
This allows testing all parts of the codebase, and can provide richer error detection and reporting. However, the tests are not as readable or portable as the XML tests.

  • To run the unit tests in a module (jts-core):

      mvn test -pl modules/core

XML Tests

JTS provides a code-independent, declarative XML-based format for expressing geometric functional tests.
This format has the following advantages:

  • allows encoding large geometries
  • provides geometric test cases in a reusable way
  • easily consumed by external tools such as the JTS TestBuilder or other geometry libraries (e.g. GEOS)
  • allows geometric tests to be used with other operation implementations, for testing or comparison purposes

This format should be used for tests which involve large geometries, or which express fundamental geometric semantics of the JTS library.