Please note that this version of VerveineJ is compatible with Moose-10 (and possibly later). It is based on a meta-model that does not match the one of previous versions of Moose, hence, they are not able to load the models generated by this VerveineJ.
A Java to JSON/MSE importer.
Based on JDT, it parses java code to export it in the MSE or JSON formats used by the Moose data analysis platform.
There is a docker version at: https://github.com/Evref-BL/VerveineJ-Docker) (thanks to Benoît Verhaeghe for this). If you already have docker the summary is:
docker run -v <full/path/toSource>:/src [-v <full/path/toDependency>:/dependency] ghcr.io/evref-bl/verveinej:latest <verveineJOption>The resulting model file (.json or .mse) is in <full/path/toSource>.
Otherwise, after the installing VerveineJ locally, the simplest command is
./verveinej.sh <verveineJOption> <java-source-directory>It will create an output.mse (JSON format also available) file with the model extracted from the <java-source-directory>
To see what other options are available:
verveinej.sh -hYou only have to clone this project and then run verveineJ
# https
git clone https://github.com/moosetechnology/VerveineJ.git
# ssh
git clone [email protected]:moosetechnology/VerveineJ.gitTo test the project, remember that you must disable the assert by removing (or not using) the -ea parameter.
You also need to run tests one by one (fork method in IntelliJ).
You can also use Ant or the pre-created IntelliJ build
ant junitSometimes, because of the JVM optimization, stacktraces are not fully displayed. To force JVM keep the full stacktrace, you can use the following vm options:
-XX:-OmitStackTraceInFastThrow -XintThis will make the execution super slow, so keep this option for debug purpose only.