Extime fetches flights information provided by Avinor, the company operating most Norwegian airports, and converts them into NeTEx-compliant timetables.
Avinor publishes flights information through its flight data web service portal, using a custom XML format.
These datasets are downloaded, combined and converted into NeTEx timetables compliant with the Nordic NeTEx Profile. Ultimately the flight timetables are injected into the Norwegian National Journey Planner and used to provide multi-modal journey planning.
Extime makes use of the following dataset provided by Avinor:
- Flights timetables: https://asrv.avinor.no/XmlFeed/v1.0
Extimes converts the flight data into a set of NeTEx documents packaged as a zip file following the Nordic NeTEx Profile delivery format.
Extime extracts and converts only a subset of the flights published by Avinor. Flight are filtered through a whitelist of airports and a whitelist of airlines.
Extime generates NeTEx timetables that reference airports as NeTEx StopPlaces and Quays. These StopPlaces and Quays should be declared in the Norwegian Stop Place Register.
Airports and airlines are configured in:
-
no.rutebanken.extime.model.AirportIATA: Airports whitelist -
no.rutebanken.extime.model.AirlineIATA: Airlines whitelist -
netex-static-data.yml: Metadata for airports and airlines
General application configuration parameters are set in the Spring Boot configuration file application.properties
Main parameters:
| Parameter name | Description |
|---|---|
| avinor.timetable.feed.endpoint | Flights timetables endpoint |
| avinor.timetable.scheduler.consumer | Configuration of the frequency of the Flight information import |
| avinor.timetable.period.forward | Time window for which future flight data are imported |
| avinor.timetable.period.back | Time window for which past flight data are imported |
- For each whitelisted airport, Extime sends a query to the Avinor REST API and retrieves all passenger flights over a given time period.
- The API returns a list of Flight objects that represent either a departure or an arrival at a given airport.
- The arrival and departure of a given flight are matched by their unique ID and mapped to a FlightLeg object.
- A heuristic searches among all FlightLegs those that are part of a multi-leg flight:
- they share the same airline and flight number,
- a given leg arrives at the same airport as the one the next leg departs from,
- the layover between the two legs is less than 3 hours.
- Single-leg and multi-leg flights are mapped to ScheduledFlight objects.
- ScheduledFlights are grouped together to build NeTEx ServiceJourneys, JourneyPatterns, Routes and Lines
In the production environment, Extime uploads the generated NeTEx archive in a common storage area on Google Storage.
Once the archive is uploaded, Extime notifies Marduk that a new delivery is available by publishing a message on a Google PubSub topic.
Access to Google Storage and Google PubSub is configured through the properties blobstore.gcs.* and spring.cloud.gcp.pubsub.* respectively.
In a development environment:
- Google Storage can be substituted by the local file system. Use the following properties:
spring.profiles.active=local-disk-blobstore
blobstore.local.folder=/path/to/local/directory
- Google PubSub can be substituted by Google PubSub emulator. Use the following property:
spring.cloud.gcp.pubsub.emulatorHost=localhost:8085
Use the following properties to capture the XML data sent by the Avinor REST API:
avinor.timetable.dump.output=true
avinor.timetable.dump.output.path=/home/user/extime/data/dump
Use the following properties to use the dumped data as input:
avinor.timetable.dump.input=true
avinor.timetable.dump.input.path=/home/user/extime/data/input
Readiness and liveness probes are provided by Spring Boot Actuator (/actuator/health/readiness, /actuator/health/liveness)