Yes, you need to apply some algorith, that can solve the problem for you.
You can use this great libray:
to solve this part of the problem.
As to the way of storing your data, you need to define vertices (the photos) and edges between vertices like (photo A-photoB), (photo A-photo C), and so on.
You have to recover that info from the database and load the corresponding structure in quickgraph and let it find the path for you.
Here you have extensive docs and samples:
For something similar to this I used:
- MyEdges class, which implements
IEdge<T> (T should be you photo ID type, int or whatever is) - represents the edge between to photos (places)
- Graph class, which inherits
AdjacencyGraph<T,MyRelation>. You load it with the available MyEdges (this is directed graph)
- PathFinder algorithm class: I inherited from
FloydWarshallAllShortestPathAlgorithm<T, MyRelation>
Then you have to:
- Create the edges (i.e. read them from the DB)
- Instance a Graph class, and add all edges to it
- Use the PathFinder constructor, using the graph as parameter. This finds the paths.
This algorithm lets you specify to which photos you can go from a given photo (edges), supposes that the distance is similar between them, but you have to define all the routes (from A to B, from B to A and so on). That's the "un-weighted" part of the OP. If your case differs you'll have to read the docs.
You can implement an UnDirected graph if you prefer that adding A To B also adds B to A. It can spare some lines of code, but I usually prefer adding all the possibilities myself. Its easier to think "from the Library I can go to aisle A and aisle B. From aisle B to Library and Laboratory", and so on, that trying to think of all edges.
You can create two tables in a database:
- Photos (with IDs)
- Path (IdFrom and IdTo)
This is easy to maintain and implement.
