parsimony "add partial" produces unwanted structure

[westram]

When adding partial sequences using ARB_PARSIMONY each is placed next to one selected full sequence. When multiple partials are added to the same full sequence, they are grouped together in one subtree and that subtree is placed as "brother" of the full sequence.

All inner branch lengths of that subtree should be zero and only leaf branches should have a length (=the "partial distance" between full and partial sequence).

This works for 2 or fewer partial species and fails for 3 or more (see attached screenshot).

[westram]

Shows wrong placed partials (next to RhoPy162 and RhoOp626)

[westram]

by [12489]

Branchlength was not restored at insertion point, but at brother of full sequence. These two were the same for first and second partial added.

[westram]

result after fix

[westram]

mark arb6.0.x hotfixes