Ignore UnrestrictedPublicTransportAreas that do not contain regular stops

[vpaturet]

Summary

This PR fixes #4918: The OTP graph builder should ignore FlexibleStopPlaces of type UnrestrictedPublicTransportAreas when they do not contain any stop.
The FlexibleStopPlace is dropped and an issue is logged in the issue store.
Additionally a null-check is added in the GroupStop constructor to prevent the creation of a GroupStop without a centroid (that triggers NullPointerExceptions downstream in the code)

Issue

Fixes #4918

Unit tests

Added unit test

Documentation

No

[codecov]

Codecov Report

Patch coverage: 100.00% and project coverage change: +0.07 🎉

Comparison is base (cc411e6) 62.95% compared to head (bb2307a) 63.03%.

❗ Current head bb2307a differs from pull request most recent head 9df5ed0. Consider uploading reports for the commit 9df5ed0 to get more accurate results

Additional details and impacted files

@@              Coverage Diff              @@
##             dev-2.x    #4919      +/-   ##
=============================================
+ Coverage      62.95%   63.03%   +0.07%     
- Complexity     13276    13310      +34     
=============================================
  Files           1663     1663              
  Lines          66341    66399      +58     
  Branches        7221     7232      +11     
=============================================
+ Hits           41768    41855      +87     
+ Misses         22194    22168      -26     
+ Partials        2379     2376       -3     

Impacted Files	Coverage Δ
...opentripplanner/netex/mapping/FlexStopsMapper.java	86.11% <100.00%> (+1.98%)	⬆️
...etex/mapping/calendar/DayTypeAssignmentMapper.java	86.95% <0.00%> (-4.12%)	⬇️
...ipplanner/netex/mapping/GroupOfStationsMapper.java	15.38% <0.00%> (-2.80%)	⬇️
...entripplanner/netex/mapping/TripPatternMapper.java	70.80% <0.00%> (ø)
...org/opentripplanner/netex/mapping/RouteMapper.java	87.01% <0.00%> (+1.09%)	⬆️
...org/opentripplanner/netex/mapping/NetexMapper.java	82.23% <0.00%> (+1.89%)	⬆️
...lanner/netex/loader/parser/ServiceFrameParser.java	73.97% <0.00%> (+2.05%)	⬆️
...opentripplanner/netex/mapping/StopTimesMapper.java	66.66% <0.00%> (+3.38%)	⬆️
...in/java/org/opentripplanner/netex/NetexBundle.java	92.98% <0.00%> (+3.50%)	⬆️
...planner/netex/loader/NetexDataSourceHierarchy.java	90.69% <0.00%> (+4.65%)	⬆️
... and 5 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[leonardehrenfried]

Since I am the opposite, I'm curious: why do you prefer to prefix the test methods with test?

[vpaturet]

this is a recommendation of SonarQube.
test* is the recommended pattern.

[vpaturet]

(and for the actual reasoning: a method name should reflect what the method does. The method does not map invalid area stops, it tests what happens when invalid area stops are mapped)

[leonardehrenfried]

The second is actually a good point. I had never thought about it that way.