README.md

VerveineJ

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.

Running it

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 -h

Installation

Installation page

You only have to clone this project and then run verveineJ

# https
git clone https://github.com/moosetechnology/VerveineJ.git

# ssh
git clone git@github.com:moosetechnology/VerveineJ.git

Testing

To 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 junit

Debug advanced mode

Sometimes, 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 -Xint

This will make the execution super slow, so keep this option for debug purpose only.