Add validation to the serviceJourney Transmodel API endpoint

[eibakke]

Summary

This change adds validation to the serviceJourney endpoint in the Transmodel API by requiring that at least one of the argument lists be non-null. It also aligns the meaning of passing an empty list and an empty string as values of the arguments. More details in the issue linked below. This change does not require that there always be at least one active date set, therefore this change does not close the issue.

Issue

#6545

Unit tests

No unit tests in this part of the code. Ran locally to verify that things are working.

[codecov]

Codecov Report

Attention: Patch coverage is 9.09091% with 10 lines in your changes missing coverage. Please review.

Project coverage is 70.24%. Comparing base (c2355db) to head (36fec83).
Report is 8 commits behind head on dev-2.x.

Files with missing lines	Patch %	Lines
...anner/apis/transmodel/TransmodelGraphQLSchema.java	12.50%	7 Missing ⚠️
...entripplanner/apis/transmodel/support/GqlUtil.java	0.00%	3 Missing ⚠️

Additional details and impacted files

@@              Coverage Diff              @@
##             dev-2.x    #6559      +/-   ##
=============================================
- Coverage      70.25%   70.24%   -0.02%     
+ Complexity     18387    18386       -1     
=============================================
  Files           2088     2088              
  Lines          77425    77429       +4     
  Branches        7843     7842       -1     
=============================================
- Hits           54395    54389       -6     
- Misses         20258    20271      +13     
+ Partials        2772     2769       -3     

☔ 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.

[habrahamsson-skanetrafiken]

I think in general it's best to be strict and not have special handling for empty strings and other edge cases. Cleaning user data hides bugs in the clients where they should be fixed.

If a user says they want the service journeys with private codes [ "" ] then they should get the service journeys that has that private code, which will most probably be an empty set.

What do you think about that?

The null filtering is fine though, we probably need this since we for some reason allowed the elements to be null.

[eibakke]

This is the opposite of what @vpaturet expressed in #6545 (comment)

I tend to agree with Vincent here, that empty strings should be ignored ie. treated the same as nulls. This also matches how the mapIDsToDomain method handles empty (or even blank strings), using

OpenTripPlanner/utils/src/main/java/org/opentripplanner/utils/lang/StringUtils.java

Line 25 in e8e9a77

public static boolean hasValue(String text) {

[habrahamsson-skanetrafiken]

Sorry for the slow response here!

I must say that I disagree with treating empty strings as nulls. If we need to do it for backwards compatability I can see the point, but in this case it looks like we are making the api less strict than previously. The stricter we make the api and the less special cases we add the easier it will be to maintain.

We had a short discussion around this on the developer meeting two weeks ago, and I think the consensus was to avoid special handling for empty strings unless it's needed for compatability reasons.

[optionsome]

You have to regenerate schema.