Fix exit condition in calc_distances

[foxtran]

Probably affects #621

Before patch, one distance were skipped between atoms # N and # N-1. For single-bonded systems (like H2) it is critical since it means that they were not bounded and then a lot of artefacts happen.

[TyBalduf]

Might be missing something, but I don't see how this would affect #621. This function seems to only be used for printing a summary of atom distances at the end of a calculation, but the NaN's in the linked issue start to occur during the initial SCF.

The PR here does fix a real bug, as the outputs from the linked issue do show that the "selected distances" section isn't being printed, which should only occur if no distances were found. More generally, the summary should have been printing N(N-1)/2 distances, but it presumably has always printed one less distance.

The code change looks reasonable to me, but I don't think it resolves #621.

[foxtran]

@TyBalduf, yep. I found more bugs with H2 system, so it may not fix issue with #621

UPD: No, I fixed exactly this bug :( Just started collecting patches from scratch.