Add agencies and feed publisher to geocoding result

[leonardehrenfried]

Summary

@miles-grant-ibigroup has requested that some data be added to the sandbox geocoder result, namely the agencies and feed publisher of a result.

It also shortens the path by removing the now superfluous /routers/default part, which is in line with other PRs and described in #4052.

Unit tests

Added.

Documentation

Updated.

[codecov]

Codecov Report

Attention: 5 lines in your changes are missing coverage. Please review.

Comparison is base (eaa9160) 67.50% compared to head (a0e40e3) 67.52%.

❗ Current head a0e40e3 differs from pull request most recent head a83c2b1. Consider uploading reports for the commit a83c2b1 to get more accurate results

Files	Patch %	Lines
...opentripplanner/ext/geocoder/GeocoderResource.java	0.00%	2 Missing ⚠️
...pentripplanner/ext/geocoder/StopClusterMapper.java	87.50%	1 Missing and 1 partial ⚠️
...in/java/org/opentripplanner/apis/APIEndpoints.java	0.00%	1 Missing ⚠️

Additional details and impacted files

@@              Coverage Diff              @@
##             dev-2.x    #5634      +/-   ##
=============================================
+ Coverage      67.50%   67.52%   +0.01%     
- Complexity     16271    16296      +25     
=============================================
  Files           1887     1889       +2     
  Lines          71545    71588      +43     
  Branches        7398     7409      +11     
=============================================
+ Hits           48300    48341      +41     
- Misses         20739    20748       +9     
+ Partials        2506     2499       -7     

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

[binh-dam-ibigroup]

Could you describe a case where multiple agencies would be returned?

[binh-dam-ibigroup]

In the Atlanta region, for the stop cluster formed by stops with code 600011 (from 3 different feeds for Cobb, Gwinnett, and Xpress), the logic as written returns one stop as documented, however only one agency is listed. Is that intended?

[binh-dam-ibigroup]

I see now. See my comment in GeocoderAPI.md.

[binh-dam-ibigroup]

After consulting with @miles-grant-ibigroup, this comment https://github.com/opentripplanner/OpenTripPlanner/pull/5634/files#r1463811604 needs to be addressed, so that all agencies operating stops in a stop cluster are listed.

[leonardehrenfried]

After consulting with @miles-grant-ibigroup, this comment https://github.com/opentripplanner/OpenTripPlanner/pull/5634/files#r1463811604 needs to be addressed, so that all agencies operating stops in a stop cluster are listed.

The current implementation only ever returns a single stop (admittedly the end point is called "cluster") and a single ID. Which stop it is, is chosen at random and the definition of a cluster are stops . If we wanted the result to return multiple stops from multiple feeds I would have to rework the API a bit. Is this what we want, @binh-dam-ibigroup @miles-grant-ibigroup?

BTW, even in a single feed a stop can be used by several agencies. There is no one-to-one mapping between stop and agency, the relationship is stop->route->agency.

[leonardehrenfried]

If there was a single, integrated feed in Atlanta that would contain these three agencies and a station/stop hierarchy, this would work automatically.

[leonardehrenfried]

BTW, the current Pelias geocoder on atlrides.com shows nothing for the term "600011".

[binh-dam-ibigroup]

BTW, the current Pelias geocoder on atlrides.com shows nothing for the term "600011".

However the OTP geocode endpoint does: https://atl-rides-otp-v2.ibi-transit.com/otp/routers/default/geocode?query=600011

[binh-dam-ibigroup]

Approving per team discussion.