Export the network of the FOLLOWER relationship in gitminer

graphmlExport.groovy

/**
 * This script will create a graphml file with the follower relation
 * every user in a gitminer database.
 *
 * @author Patrick Wagstrom <patrick@wagstrom.net>
 */

/**
 * Process the traversal to save FOLLOWER relationships
 *
 * This eliminates the biography property because it takes up a ton
 * of space and we're not that concerned about it for our analysis.
 *
 * @param users: an iterable list of usrs. Must be all users in the
 *   set of you could experience some problems.
 * @param outfile: the file to save the graph to
 */
def exportFollowerSubgraph(users, outfile) {
    to = new TinkerGraph()
    System.out.println("processing users")
    for (vertex in users) {
        System.out.println("user: " + vertex["login"])
        toVertex = to.addVertex(vertex.getId())
        ElementHelper.copyProperties(vertex, toVertex)
        toVertex.setProperty('biography', null)
    }

    System.out.println("processing edges")
    for (vertex in users) {
        System.out.println("edges for user: " + vertex["login"])
        for (edge in vertex.outE("FOLLOWER")) {
            if (edge == null) continue;
            System.out.println("edge: " + vertex["login"] + " => " + edge.getInVertex()["login"])
            // System.out.println("to.v: " + to.v(vertex.getId()))
            if (to.v(edge.getInVertex().getId()) == null) {
                System.out.println("target vertex is out of set, skipping");
            } else {
                toEdge = to.addEdge(edge.getId(), to.v(vertex.getId()),
                    to.v(edge.getInVertex().getId()), edge.getLabel())
                ElementHelper.copyProperties(edge, toEdge)
            }
        }
    }

    System.out.println("Writing graphml file")
    GraphMLWriter writer = new GraphMLWriter()
    writer.outputGraph(to, new FileOutputStream("follower.graphml"))
}

/**
 * this just gets a couple of different ways that users can be
 * associated with a project.
 *
 * a future post will contain more information about how to mine these
 * relationships.
 *
 * @param repo: the vertex of the repository to start at
 */
def getAllUsersForRepo(repo) {
    watchers = repo.in("REPO_WATCHED").toSet()
    collaborators = repo.out("REPO_COLLABORATOR").toSet()
    contributors = repo.out("REPO_CONTRIBUTOR").toSet() + \
                   repo.in("REPO_OWNER").dedup().toSet()

    return watchers + collaborators + contributors
}

// get every user related to gremlin
gremlin = g.idx('repo-idx').get('reponame','tinkerpop/gremlin').next()
// println "Getting all users..."
users = getAllUsersForRepo(gremlin)

// save the graph to followers.graphml
exportFollowerSubgraph(users, "followers.graphml")