Enrich results with origin/destination

[xoolive]

Feature request by email from Davide Falanga

We've noticed that queries like the one below do not return some important features for our work, such as 'origin', 'destination', and 'aircraft_type'. In the past, at least the first two features were returned when using queries with more parameters (e.g., including the parameter 'airport').

trajectories = opensky.history(
    start=date,
    stop=date + timedelta(hours=2),
    bounds=(west, south, east, north)
)

Is there a way to retrieve these features in a query like the one shown above?

[xoolive]

Few thoughts:

• it is easy to get those columns when you build a query using the flight table (and usually, filter flights based on origin/destination), because you first get the flight list, then go with the join query to download the state vectors
• it is easy to write an extra query in the flight table based on the results of the one written above, maybe suboptimal?
• if we can write a proper query to merge the results of the query above with the flight table:
  • how to write it efficiently?
  • what kind of arguments should we add to opensky.history to trigger that behaviour?

[espinielli]

See comment on issue #87 for a solution to the request to retrieve the (estimated) aerodrome of departure and destination like in the old API.

A possible automatic inclusion of origin and destination is to insert FlightsData4.estdepartureairport and FlightsData4.estarrivalairport columns in the selected_columns whenever any of departure_airport, arrival_airport, airport is not None in history(), and rename the returned estdepartureairport and estarrivalairport respectively to origin and destination

[xoolive]

See comment on issue #87 for a solution to the request to retrieve the (estimated) aerodrome of departure and destination like in the old API.

This works in your example in #87 because you have airport="LSZH".

In the current use case the filter is on the bounding box, so there is no call to the ˋflights_data4ˋ table necessary and the extra columns have no reason to exist.