struts ignoring pbonds and accumulating inappropriately

[Elaine Meng]

The "struts" command calculation is not paying attention to displayed pseudobonds, and besides that, seems to be generating far too many struts in subsequent rounds of calculations even when I close the struts model in between.

In Chimera,
open 1zik; struts @ca
...gives 8 struts, and behaves as expected after I add two pseudobonds, e.g.
close 1; dist :12@ca; dist :19@ca; struts @ca
...gives 6 struts.

In ChimeraX,
open 1zik; struts @ca
...gives 9 struts (that discrepancy doesn't bother me, the later stuff does), but after adding two pseudobonds (file attached)
close #2; open ~/Desktop struts-test.pb; struts @ca
...gives 40 struts. Thinking maybe my pb file added atoms, I then tried
close #3; struts #1@ca
...but that gives 63 struts!

If I start over completely and open the pb file before any struts calculation, it does not pay attention to them and still gives 9 struts:
open 1zik; open ~/Desktop struts-test.pb; struts @ca

[Tom Goddard]

These strut problems appear to be caused by hidden atoms of ribbons. If all atoms are displayed then the examples all work correctly.
Investigating how hidden atoms are not handled correctly.

[Tom Goddard]

Fixed.

Main problem was that backbone atoms hidden by ribbon were not being considered. This came about because the struts command told the ribbon command not to hide the atoms so that a strut to a ribbon would include the backbone atom the strut ends on. But then the strut command run again will only consider displayed or hidden atoms and now the ribbon backbone atoms are mostly undisplayed and not hidden so connectivity is lost along the backbone and tons of struts are made. I didn't see any good remedy for this so I changed struts so it does not try to unhide ribbon backbone atoms. The drawback of that is that the strut cylinder end is not capped by a sphere, instead it is a flat cap which can be ugly ending on a ribbon. Usually those ends are not visible because the ribbon is fattened so we'll live with that.

Another problem was that pseudobonds to hidden atoms were not being considered. Fixed that.