Future improvements

[hasanbalci]

1. While solving pd10141 (All process nodes should have at least one input and at least one output), maybe we can check whether there is a orphaned EPN or not. If there is, we can suggest that EPN as the default choice. In this way, the error pd10131 (EPNs should not be orphaned (i.e. they must be associated with at least one arc)) caused by the orphaned EPN will also be solved automatically.

2. Again for the error pd10141, if a process node has an input as EPN but doesn't have an output, we suggest nearby EPNs as alternatives, but the input EPN is selected as the default option (probably because it is the nearest?). But it is unlikely that the input and output of a process will be the same EPN (even though it is possible in terms of validation rules). Maybe we can keep that EPN in the alternatives, but not select as the default option if there is any other option.

3. In this example new_file_9.nwt.txt, we get pd10109 and pd10110 errors which indicate a symmetric error (like in the case of pd10101 and pd10102) and it is actually caused by the reverse connection of the modulation arc to its source and target nodes. Here, while swapping the source and target nodes resolves both errors at the same time, we only change the source to the closest EPN and leave the target with an error.
   In case of this symmetric errors pd10109 and pd10110, we can apply the same approach with pd10101 and pd10102 and first try to swap source and target if the same edge is causing both errors.

4. Related to item 3, (pd10101 and pd10102), (pd10109 and pd10110) and (pd10127 and pd10128) are symmetric errors and swap operation can be tried first and if doesn't resolve the issue, we can present options to user with nearby nodes. We currently apply swapping in the first one and offer options in the other two.