Integration tests for Geofabrik Overpass API

[nataliejschultz]

Addressing e-mission/e-mission-docs#1036

As of now, there are three tests in the TestOverpass.py module:

1. test_overpass, which compares a call to the free version of overpass with a call to the geofabrik version
2. test_get_stops_near, which passes a set of coordinates (as a list inside a dictionary) to get_stops_near in match_stops.py.
3. test_get_predicted_transit_mode, which passes two sets of coordinates to get_predicted_transit_mode in match_stops.py.

[nataliejschultz]

Currently, tests 2 and 3 only call the free version of the API to make calls. There are a few ways that come to mind to change this:

1. Add hard-coded query to geofabrik as seen in test 1 and compare with the free call (not preferred)
2. Add a similar block of code to match_stops.py as we did for nominatim.py using environment variables:

try:
    NOMINATIM_QUERY_URL = os.environ.get("NOMINATIM_QUERY_URL")
    logging.info(f"NOMINATIM_QUERY_URL: {NOMINATIM_QUERY_URL}")
    print("Nominatim Query URL Configured:", NOMINATIM_QUERY_URL)

    if NOMINATIM_QUERY_URL is None:
        raise Exception("Nominatim query url not configured")
except:
    print("Nominatim URL not configured, place decoding must happen on the client")

3. Add a file called overpass_server.json with the geofabrik query URL and work with the current method of setting the query URL in match_stops.py

try:
    config_file = open('conf/net/ext_service/overpass_server.json')
except:
    print("overpass not configured, falling back to default overleaf.de")
    config_file = open('conf/net/ext_service/overpass_server.json.sample')
config_data = json.load(config_file)
url = config_data["url"]

I think the answer is a combination of 2 and 3, but I'd like some feedback from @shankari on my tests before deciding what to do

[shankari]

@nataliejschultz high-level question: why do we need to keep the configuration in a file? There is a single variable in the file.
Recall that for the nominatim use case, we switched the config from a file to an environment variable. I would suggest using the same pattern here.

[nataliejschultz]

Coming back to this as I wait for review on other PRs.
The overpass url is now based on setting the OVERPASS_KEY environment variable in match_stops.py (rather than sample file), to be consistent with current direction of codebase:

try:
    OVERPASS_KEY = os.environ.get("OVERPASS_KEY")
    url = 'https://overpass.geofabrik.de/' + OVERPASS_KEY + '/'
    print("overpass configured")
except:
    print("overpass not configured, falling back to public overpass api")
    url = "https://lz4.overpass-api.de/"

I made sure that using unset OVERPASS_KEY in terminal triggered the "overpass not configured" print, and it did:

This readies the code to use the overpass key for calls in both the tests and any usage of match_stops.py.

Tests passing on GitHub Actions and locally: