reproin: might want to change logic for dups to actually use datetime not seq number

since unlikely to ever become part of the official bids validator, we might need "our own" validation helper to detect such already accomplished problems.

Here is a script from @jungheejung she used to find those problematic cases where __dup functional were longer than non-dups, indicating that non-dup is likely a canceled run: https://github.com/spatialtopology/spacetop-prep/blob/master/spacetop_prep/datalad/remove_dups_anatfmap.sh

also in https://github.com/spatialtopology/spacetop-prep/issues/57#issuecomment-1961651646 I had used following oneliner to sort all sequences by time and not index to spot the problem:

[bids@rolando A004936] > for f in 0*/; do dcmdump +P 0008,0032  $f/0000011.dcm \
 | sed -e 's,\[,,g' -e 's, *#.*,,g' | sort -n -k 4 | tr '\n' ' '; echo  "$f"; done | sort -n -k 3

nipy / heudiconv

reproin: might want to change logic for dups to actually use datetime not seq number #740