fix: panic removing InPorts in MultiPortGraph::set_num_ports

[acl-cqc]

See test case. The problem is that iterating through the dropped_ports reads self.multiport using old indices, after self.multiport was mutated/renumbered by the callback earlier. (Thus, if the first OutPort was a multiport, removing the last InPort can end up reading a false positive.)

I've tried to see if I can break it by adding inports too, but in such cases all the ports get moved to a new "block" so there is no problem (hence, a more thorough "fix" to do all reading of self.multiport before any writing to it, including the reading+writing performed by self.multiport.swap, does not appear necessary).

[codecov]

Codecov Report

Attention: Patch coverage is 80.00000% with 4 lines in your changes missing coverage. Please review.

Project coverage is 83.60%. Comparing base (c9e7483) to head (a648a79).
Report is 1 commits behind head on main.

Files with missing lines	Patch %	Lines
src/multiportgraph.rs	80.00%	2 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #191      +/-   ##
==========================================
+ Coverage   83.56%   83.60%   +0.04%     
==========================================
  Files          24       24              
  Lines        6347     6363      +16     
  Branches     6347     6363      +16     
==========================================
+ Hits         5304     5320      +16     
+ Misses        969      967       -2     
- Partials       74       76       +2     

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[aborgna-q]

Ouch. Good catch!