Configuring r:trån

Transformation

The transformation of models is usually dependent on a set of parameters. r:trån can be configured by adding a script named configuration.kts into the directory of your source model:

import io.rtron.main.project.configuration.configure

configure {

    opendrive2roadspaces {
        tolerance = 1E-7
        attributesPrefix = "opendrive_"
        crsEpsg = 32632
    }

    roadspaces2citygml {
        discretizationStepSize = 0.5
        sweepDiscretizationStepSize = 0.3
        circleSlices = 12
        flattenGenericAttributeSets = true
    }
}

This script is an internal DSL of Kotlin which guarantees type-safety. Moreover, a DSL is suitable for describing transformation recipes and complex configurations, such as model mapping rules.

If you want to edit the configuration, IntelliJ is recommended as it has the best support for Kotlin. An example configuration script can be found at rtron-main/src/main/kotlin/io/rtron/main/project/configuration/examples/configuration.kts.

Batch Processing

Usually the transformation parameters are similar per project. To simplify parametrization for batch processing transformations, r:trån supports nested configurations:

tree ./opendrive-datasets 
  ./opendrive-datasets
  ├── asam.net
  │   ├── configuration.kts
  │   ├── Ex_Line-Spiral-Arc.xodr
  │   └── UC_ParamPoly3.xodr
  ├── configuration.kts
  └── opendrive.org
      ├── configuration.kts
      └── CrossingComplex8Course.xodr

A parameter in ./opendrive-datasets/asam.net/configuration.kts will overwrite the equally named parameter in ./opendrive-datasets/configuration.kts. All parameters provided in ./opendrive-datasets/configuration.kts will also apply to the subdirectories, in case there is no parameter overwriting. Default values are used, if no parameters are provided.

Java Virtual Machine (JVM)

If larger OpenDRIVE datasets are to be transformed, the JVM should be provided with enough heap space:

java -Xmx50G -jar rtron-*.jar ./opendrive-datasets ./citygml-datasets