longest continuous segment algorithm

an attempt to code lcs as described here. the signature of the only exported method is (protein1, protein2, cut), where:

• proteinN, N in [1,2], is an array of protein CA atoms positions:

[
  [x1, y1, z1],
  [x2, y1, z2],
  ...
]

• cut is angstrom rmsd cut for lcs run

return value has the following format:

{
  subs: [ { tgt: [], exp: [] }],
  stats: []
}

where subs is an array of all matched segments in the second protein (model, tgt) and the first protein (exp). tgt and exp arrays in each record are zero-based references to residues in second and first proteins, respectively. stats is an array indexed by zero-based residue indices of first protein, where the value at i is the length of the longest segment in second protein where i-th residue of the first protein, while being matched - probably along with other residues, had the longest match.