Changed graph-building logic to use full project names.

Maven and Ivy Projects that are configured to build their modules as separate jobs can sometimes have modules with identical names in different projects. Using project.getName() meant that these modules were treated as the same graph node.
Using project.getFullName() ensures a unique string that correctly identifies these sub-projects.  Also, using project.getFullDisplayName() for the actual graph output results in a more readable output.

src/main/java/hudson/plugins/depgraph_view/AbstractDependencyGraphAction.java

@@ -77,7 +77,7 @@ public int compare(Dependency o1, Dependency o2) {
     private static final Comparator<AbstractProject<?,?>> PROJECT_COMPARATOR = new Comparator<AbstractProject<?,?>>() {
         @Override
         public int compare(AbstractProject<?,?> o1, AbstractProject<?,?> o2) {
-            return o1.getName().compareTo(o2.getName());
+            return o1.getFullName().compareTo(o2.getFullName());
         }
     };
 
@@ -161,14 +161,14 @@ public String generateDotText(Set<AbstractProject<?,?>> projects, Set<Dependency
     }
 
     private String projectToNodeString(AbstractProject<?, ?> proj) {
-        return escapeString(proj.getName()) +
+        return escapeString(proj.getFullDisplayName()) +
                 " [href=" +
                 escapeString(Hudson.getInstance().getRootUrlFromRequest() + proj.getUrl()) + "]";
     }
 
     private String dependencyToEdgeString(Dependency dep) {
-        return escapeString(dep.getUpstreamProject().getName()) + " -> " +
-                escapeString(dep.getDownstreamProject().getName());
+        return escapeString(dep.getUpstreamProject().getFullDisplayName()) + " -> " +
+                escapeString(dep.getDownstreamProject().getFullDisplayName());
     }
 
     private String escapeString(String toEscape) {