Adding 1D benchmark with conflicts

[Mommessc]

New PR following #1

[Mommessc]

@surdykaw some new discussions before integrating the 1D BPPC benchmarks.

I think it would be better for the BPPC_dataset archive to keep the same organisation as the BPPLIB Falkenauer instances: two folders for the "uniform" and the "triplet" types. And in each folder there could be subfolders for each variant (parameters for the p-generator and graph density).
In each subfolder there would be the 80 instances (20 x 4 different number of items) in the VBP format augmented with the conflicts.

For example:

- BPPC_dataset
    - Falkenauer_T
        - A01_B03_seed42_20percent
            - 80 instances of the form Falkenauer_t60_00.vbp
        - A01_B03_seed42_40percent
            - 80 instances of the form Falkenauer_t60_00.vbp
        ...
    - Falkenauer_U
        - A01_B03_seed42_20percent
            - 80 instances of the form Falkenauer_u120_00.vbp
        ...

Regarding the instance format with conflits, it should be specified somewhere what the list of conflicts means, and that it is a "triangular adjacency matrix" (not sure this is the correct term).
Also, I think it should be explicit that the instance file are using conflicts, so maybe using the file format vbpc instead? (I am not sure, because we could also make Vectorpack_cpp able to recognize if there is a conflict list for each item, or discard the conflict list if the --with-conflicts option is not present).