Experiments: Mining, navigating and enriching software ecosystem dependency graphs

This archive contains the experiences of the chapter "Mining, navigating and enriching software ecosystem dependency graphs".It consists of four files: goblinMetrics.zip This file contains the Python script used to retrieve the information from Table 3.4 "Quantitative analysis of Goblin-Weaver metrics in Maven ecosystem".To run the script, please use the "with_metrics_goblin_maven_30_08_24.dump" Neo4j dataset available here. goblinWeaver-1.0.0.jar Version of the Goblin-Weaver used for the experiments.More information about its use here. osvData.zip External database containing CVEs.This file allows you to be at the same level as when running the experiments. mavenDatasetExperiences.zip Java project containing all experiments except the one in the goblinMetrics.zip file.To re-execute the experiments, follow this steps: Dowload the Neo4j dump "goblin_maven_30_08_24.dump" here. Import the dump on your neo4j application. Unzip the osvData.zip archive. Make sure the weaver jar is in the same folder as the OSV folder, and run (by changing the password and others according to your configuration):java -Dneo4jUri="bolt://localhost:7687/" -Dneo4jUser="neo4j" -Dneo4jPassword="password" -jar goblinWeaver-2.1.0.jar noUpdate Unzip the mavenDatasetExperiences.zip archive, go to the extracted folder and run "mvn clean install exec:java". Note that the ecosystem experiment consumes a lot of memory, so you may need to increase the Java heaps space of the database, the weaver and the project containing the experiments. For our part, we allocated for each 20GB (we have not done an optimal search). It's also possible to run the experiments separately by commenting on the unwanted ones in main.java file. Note that in the chapter, the “EXPERIENCE UPDATE” experiment was run with an older version of the database detailed here.