Fix SIRI ET ExtraJourney trips not appearing in departure/arrival boards

[t2gran]

Summary

This PR fixes a bug where extra journeys (GTFS ADDED trips / NeTEx ExtraJourney) were missing from arrival/departure boards. The root cause was that TimetableSnapshot.addPatternToIndex() was incorrectly checking if the stop pattern was modified in real-time, rather than checking if the pattern was created in real-time. This PR introduces a clearer distinction between real-time patterns with modified stop patterns (rerouted trips) and real-time patterns for completely new trips (added trips/extra journeys).

Changes

• Two boolean flags distinguish different real-time pattern types:
  • realTimeTripPattern: True for any pattern created by real-time updates (rerouted OR added)
  • stopPatternChangedInRealTime: True only when a scheduled trip's stop pattern is modified. This is exactly
    the same as the previous createdByRealtimeUpdater flag.
• TimetableSnapshot.addPatternToIndex() now uses isRealTimeTripPattern() to include all real-time patterns
• Builder methods clarify intent:
  • withRealTimeStopPatternChanged(): For rerouted trips with modified stop patterns
  • withRealTimeAddedTrip(): For completely new trips (GTFS ADDED / NeTEx ExtraJourney)
  • : For scheduled trip patterns.

Core changes

• TripPattern.java: Added dual flags, new methods, improved JavaDoc
• TripPatternBuilder.java: New builder methods with clear semantics
• TimetableSnapshot.java: Fixed index update logic (line 600)

Needs more investigation

• The use of TripPattern#isStopPatternChangedInRealTime() in TransferIndexGenerator and RealTimeRaptorTransitDataUpdater should be investigated. It is probably better to use the more generic isRealTimeTripPattern() in these places, and than add logic for handling ADDED trips (even if that logic is just to ignore these cases). The code as is, is easy to misunderstand.

Build on top of #7130

Issue

Fixes #7008

Testing

• ✅ Unit tests are updated
• ✅ Test coverage includes both added trips and rerouted trips

Manual Testing

I have manually tested the case documented in the issue, before and after applying the fix.

Documentation

• ✅ JavaDoc updated for all modified public methods
• ✅ Added comprehensive documentation explaining the distinction between realTimeTripPattern and stopPatternChangedInRealTime
• ✅ Added TODO comments highlighting areas for future improvement

Changelog

• ✅ Should be included in changelog: This fixes a user-visible bug where added trips (GTFS ADDED / NeTEx ExtraJourney) were missing from arrival/departure boards.

Bumping the serialization version id

• ✅ Needed because of the new flag in TripPattern

[codecov]

Codecov Report

❌ Patch coverage is 88.46154% with 3 lines in your changes missing coverage. Please review.
✅ Project coverage is 72.06%. Comparing base (2891fc2) to head (3509ce3).
⚠️ Report is 418 commits behind head on dev-2.x.

Files with missing lines	Patch %	Lines
...tripplanner/transit/model/network/TripPattern.java	71.42%	1 Missing and 1 partial ⚠️
...nsit/mappers/RealTimeRaptorTransitDataUpdater.java	50.00%	0 Missing and 1 partial ⚠️

Additional details and impacted files

@@              Coverage Diff              @@
##             dev-2.x    #7052      +/-   ##
=============================================
- Coverage      72.09%   72.06%   -0.03%     
- Complexity     20694    20910     +216     
=============================================
  Files           2244     2282      +38     
  Lines          83899    84466     +567     
  Branches        8347     8434      +87     
=============================================
+ Hits           60488    60873     +385     
- Misses         20482    20623     +141     
- Partials        2929     2970      +41     

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

🚀 New features to boost your workflow:

• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[leonardehrenfried]

You have to resolve the conflicts, I'm afraid.

[habrahamsson-skanetrafiken]

I think guard statements should go at the top or next to where we set the parameters. Otherwise it is easy to miss it when reading the code.

[t2gran]

I agree, but for constructors we often do validation it at the end, or where the assignment happens. Doing it at the end has the power that you can do it on the members - not the arguments. So, if there is logic in the constructor you validate the result. In short, syntactic validation (preconditions) of input fields should be done first (or part of assignment this.a = requireNonNull(a)), semantic validation last(postconditions). This constructor follow this.

[vpaturet]

Should it be withRealTimeStopPatternChanged() in the case of a modified trip and withRealTimeAddedTrip() in the case of an added trip?

[t2gran]

This is not changed, it might be wrong - but in that case it should be fixed in a separate PR. The old code:

tripPattern = TripPattern.of(id)
        :
        .withCreatedByRealtimeUpdater(true)

The withCreatedByRealtimeUpdater(true) was set for all TripPatterns where the stop-pattern was modified. But, NOT if the pattern was new.