Return only sample flows in reachableByFlows + optimizations

[fabsx00]

Prep work for upcoming article on data flow.

It is computationally prohibitive to enumerate all flows, however, determining whether a flow exists and returning a sample can be done efficiently. This PR modifies reachableByFlows to accomplish this, return the flow of maximum length. If multiple flows of maximum length exist, we sort by node-ids to ensure that we obtain reproducible.

• Return only sample flows in data flow calculation
• Use vectors as opposed to lists to store path elements and results
• Deduplicate results before writing them to the result table
• Do not store partial results if this is the last analysis stage