BRO Conepenetrationtest Converter

Contributing
Documentation
Requirements
Code Explanation
Links

Contributing

This project will not be released and neither be pushed (as Maven artefact) to a SonaType (Nexus) repository (unless there's a specific demand for this). It will remain in a version < 1.0.

However, if you do want to contribute, this project follows the Fork & Pull development approach. To get started just fork this repository to your GitHub account and create a new topic branch for each change. Please start commits with the GitHub issue number #<nr>. Once you are done with your change, submit a pull request against the CPT_GEF_CONVERTER repo.

Documentation

Mapping documentation comes in 2 documents on this site:

The Handreiking (starters guide): explains the basic principles and should be read prior to read the mapping itself.
The Mapping: explains in full detail all fields on the GEF site, the rules to which they are checked, and the resulting IMBRO / IMBRO/A xml fields. Please note that rules concerning parties and geometry are not implemented since they require on-line services.

Requirements

Java8 (JDK) is required on the path.
For building the code, Maven version 3 or higher is required on the path.

Code explanation

Building the source code

The source code can be built by the maven clean install command on a Linux / Unix shell or in a Windows command window.

Generation of the XML parser

The gef-jaxb sub module contains the XSD files locally (good practice). A catalog file is (src/main/resources/META-INF/catalog.xml) used to 'connect' the (public) URL of the schema locations to the local files. The binding file (src/main/resources/META-INF/binding.xjb) determines how the xsd (src/main/resources/schema) is mapped to the generated java classes. 2 classes are put into place manually to cope with observations OGC observation and measurements. Finally, the maven-jaxb2-plugin takes care of generating the java classes.

Implementation of the mapper

The implementation consists of several packages:

package nl.bro.cpt.gef.logic is responsible for searching a file list for cone penetration test gef files. It relates them to dissipation tests. Validation is performed on the consistency (e.g. all referred dissipation tests are present and are bound - not orphanaged - to a cone penetration test). This package supports bulk handling of gef files.
package nl.bro.cpt.gef.logic is responsible for parsing the actual GEF file. It uses an antlr definition file to generate the parser code. Initially the authors gauged that it was possible to capture the GEF standard in its entirety in a GefLexer.g4 and GefParser.g4. However, the GEF format proved to be too unstructured to do this. In the end, a minimal version has been put in place and the parser concept is used to populate the GEF data-transfer-objects.
package nl.bro.cpt.gef.dto contains the the GEF model files as so called data-transfer-objects (DTO's). The term DTO is a bit misleading because it suggests that data is actually transferred. However their only purpose is an intermediate format. This intermediate format is validated against the business rules of the BRO as far as possible. Feedback is given to the user in GEF terminology. Bean validation JSR303 is used to check the business rules. As implementation technology Hibernate Validation 4 is used, with some customizations to support messages based on the "Entity.Attribute" concept. "Entity.Attribute" uniquely identfies an attribute in the BRO model. These "Entity.Attribute" coincide with the "Object.property" in the DTO model and form the keys in the GefValidationMessages.properties.
package nl.bro.dto.gef.validation contains custom-written validation rules (Java annotations).
package nl.bro.dto.gef.transform contains the transformation to JAXB objects. It uses mapstruct to map the GEF DTO's to the generated JAXB DTO's. An implementation is generated for the Mapstruct annotated mappers.

Stand alone version

The stand-alone version of the GEF reader just builds a small wrapper around the gef_impl classes. It creates a so called 'fat' jar file that contains all the dependencies. Apache's commons-cli is used to implement a self explaining command line interface.

Just run the stand-alone with: java -jar convertgef.jar -h from the commandline.

BROprogramma / CPT_GEF_CONVERTER

readme