invalid bidi edges with ptline output

[behrisch]

see the HBF junction in the brunswick scenario. Maybe we should rather adapt the lines to the edge direction and not vice versa or make this configurable. This is also the reason for the error mentioned in #10732

[behrisch]

So it may happen that a reverse path is shorter than a straight one. To resolve this we want to introduce a configurable reversal penalty with the default of 30s.

[behrisch]

It was rather a connection problem so far

[behrisch]

There were several reasons here:

• input errors with wrong directions of tram edges
• input errors triggering check whether route and stop sequence are coherent before adding bidi #11612
• early applied patches for merging tram and road which disconnected the rail network