Added a new k disjoint shortest path algorithm for sparse graphs

[bkrogh]

Added a new k disjoint shortest path algorithm which is faster for the cases with sparse graphs and where the k disjoint paths from source to sink are similar to the shortest path tree rooted in the source vertex

This addresses issues: #796

Added a new algorithm which is a slightly altered version of the existing Bhandari implementation.

I welcome any comments that may improve the code quality (and improve chances of this getting merged)

---

• I read and understood https://github.com/jgrapht/jgrapht/wiki/Become-a-Contributor
• I read and understood https://github.com/jgrapht/jgrapht/wiki/How-to-make-your-first-%28code%29-contribution
• I added unit tests
• I added documentation
• I followed the Coding and Style Conventions
• I have not modified HISTORY.md or CONTRIBUTORS.md
• I ensured that the git commit message is a good one

[Toptachamann]

Hi,

This looks interesting. Cloud you please include an experimental running time comparison of both implementations. There're some examples of how we implement benchmark tests under jgrapht/jgrapht-core/java/test/test/org/jgrapht/perf/....

Please, also try to find out the cases when the previous implementation is more efficient. We have to state this as clear as possible in the Java doc so that there's no doubt for the users which one to use.

[jkinable]

What's the status of this PR?