Hi Nate,
as you already know arangodb produces json documents as an output.
As far as I read the desktop Cytoscape overwiew most of the import formats seem to be xml based. However, their web-version seems to be json based:
https://github.com/cytoscape/cytoscape.js/blob/master/test/collection-algorithms2.jsA question to the cytoscape developers would be, whether the desktop client can read the same graph data in json format as the webclient can - or if it can just dump it.
Then you could use ArangoDBs AQL to formulate the output results directly in the format the web client can understand:
( using this example graph:
https://docs.arangodb.com/3.0/AQL/Graphs/Traversals.html#example-execution )
LET g = (FOR v, e IN 1..3 OUTBOUND 'circles/A' GRAPH 'traversalGraph' LET vx = MERGE(v, {id: v._key}) LET ex = MERGE(e, {source: e._from, target: e._to}) RETURN {v : {data : vx} , e: {data: ex}}) RETURN { nodes: g[*].v, edges: g[*].e}
Btw, ArangoDB 3.1 is going to bring a new graph viewer, you can take a sneak preview of it here:
https://github.com/arangodb/example-datasets#debian-dependency-graphCheers,
Willi