Add feature switch for matching ET messages on stops

[Bartosz-Kruba]

Summary

Add feature switch for matching ET messages on stops

Issue

closes #4346

Unit tests

No new unit tests added

Documentation

Updated documentation

[codecov-commenter]

Codecov Report

Merging #4347 (2b13dc5) into dev-2.x (ee0859b) will increase coverage by 0.00%.
The diff coverage is 66.66%.

@@            Coverage Diff             @@
##             dev-2.x    #4347   +/-   ##
==========================================
  Coverage      57.54%   57.54%           
  Complexity     10933    10933           
==========================================
  Files           1441     1441           
  Lines          58255    58256    +1     
  Branches        6772     6773    +1     
==========================================
+ Hits           33520    33521    +1     
  Misses         22717    22717           
  Partials        2018     2018           

Impacted Files	Coverage Δ
...pplanner/ext/siri/SiriTimetableSnapshotSource.java	0.00% <0.00%> (ø)
...main/java/org/opentripplanner/util/OTPFeature.java	85.45% <100.00%> (+0.26%)	⬆️
...ripplanner/model/transfer/ConstrainedTransfer.java	80.00% <0.00%> (-3.34%)	⬇️
...pplanner/graph_builder/module/osm/OSMDatabase.java	76.47% <0.00%> (+0.18%)	⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us.

[hannesj]

I think it would be better with an updater parameter, rather than a feature flag.
Example from SIRI-VM: https://github.com/opentripplanner/OpenTripPlanner/blob/dev-2.x/src/main/java/org/opentripplanner/standalone/config/updaters/SiriVMUpdaterConfig.java#L16

[t2gran]

@Bartosz-Kruba We discussed this at the developer meeting, so if you do the switch to config - not OTPFeature as @hannesj suggest, then this looks ok.

[hannesj]

Is there some reason the parameter is not called fuzzyTripMatching as with SIRI-VM and GTFS-RT updaters? I think it would bee good to have consistency on these

[leonardehrenfried]

You want to run the formatter on your code.

[Bartosz-Kruba]

I've changed the name of config parameter name and did a rebase. Had to go for other solution since SiriTimetableSnapshotSource is no longer instantiated the same way as it was before.

[lassetyr]

With this approach, the updater still instantiates the SiriFuzzyTripMatcher, and passes it to the SiriTimetableSnapshotSource.

Would it be easier to handle this flag only in the updater, and set SiriFuzzyTripMatcher = null when this feature is switched off?

SiriTimetableSnapshotSource would then only need a null-check.

[Bartosz-Kruba]

With this approach, the updater still instantiates the SiriFuzzyTripMatcher, and passes it to the SiriTimetableSnapshotSource.

Would it be easier to handle this flag only in the updater, and set SiriFuzzyTripMatcher = null when this feature is switched off?

SiriTimetableSnapshotSource would then only need a null-check.

This is not so easy since SiriFuzzyTripMatcher is used for other things as well. It would require us to move around a lot of code. I would rather not do that since we are planning to refactor whole module anyway