fix: ensure edge-only vertices are tracked in ParseResult.getVertices()

When vertices appear only in edge lines (not in the 'nodes' section),
they were added to the JUNG graph by addEdge() but missing from
ParseResult.getVertices(). This caused inconsistencies where
graph.getVertexCount() != result.getVertices().size(), breaking
downstream code that relies on the vertex set for iteration, metrics,
or export.

Now edge endpoints are explicitly added to both the vertex set and
the graph before edge insertion, ensuring consistency regardless of
whether vertices are declared in the nodes section.

Author: sauravbhattacharya001

Gvisual/src/gvisual/GraphFileParser.java

@@ -164,6 +164,11 @@ public static ParseResult parse(String filePath, Predicate<String> visibleFilter
                         typeList.add(curEdge);
                     }
 
+                    // Ensure edge endpoints are tracked in the vertex set
+                    // (they may not appear in the "nodes" section)
+                    if (vertices.add(parts[1])) g.addVertex(parts[1]);
+                    if (vertices.add(parts[2])) g.addVertex(parts[2]);
+
                     // Only add to graph if this type is visible
                     if (visibleFilter.test(parts[0])) {
                         g.addEdge(curEdge, parts[1], parts[2]);