Boost Graph Library development on linux

Greetings, Looking for some c++ Boost Graph Library development on linux. Hoping to create an application that will assign unique edge identifiers along paths between all nodes in a non-directional graph. The application must take a matrix style graph input, then calculate the shortest path between all nodes. Then for all pairs of nodes, traverse the shortest path between them allocating a unique edge id number (between 1-4094) for each edge in the path. It may be possible to do the unique edge id allocation as a customer visitor such that merely calling johnson_all_pairs_shortest will do the allocations. Multiple outputs are needed, we are essentially an edge path view, and a node view: - Output showing the path and the unique id allocated along each edge of the shortest path between each pair of nodes. Output of this as a graphviz dot file for visualization. - Output showing for each node, all the input and output edge ids, and for each of these ids the node pairs for which the edge id is part of The program should allow ingestion of a simple graph definition from simple CSV file, representing the graph in matrix style notation. e.g. 'node a', 'node b', 'node c' … 'node n' on the top row, and the same names down the left hand column, then a zero (0) or integer (1-99) to represent if an edge exists and its weight. The column and row names like 'node a' will allow the creator of the CSV file to change the names of the nodes in the output. Example file attached, but obviously I'll be expecting you to test with larger graphs. Code must be demonstrated to work on a centos5.5 system with all the development libraries required. I will supply the system for this demonstration. Kind regards, Dave