Test the refresh_departures service.

Test user input resulting in multiple matches, requiring selection.

Test user input resulting in a single unique match.

Test that all translation_keys used in sensors are in strings.json.

Test the set_paused service.

Verify specific formatting fix in limitations.md.

Test that the main sensor publishes consolidated station messages.

Strictly validate schema compliance against March 2026 hassfest rules.

Test that all strings in strings.json are present in translations.

Regression: AVV buses use 'line' instead of 'train'. Before the fix, dep.get("train") returned None for every AVV entry, causing the `no

Regression: entries filtered early in the main loop never get departure_timestamp set. Before the fix they all produced history key "<tr

Regression: raw API cancellation field names (cancelled / isCancelled) must be recognised even when entries were filtered before the main loo

Verify that using AND logic with one station still allows server-side optimization.

Test that AND logic excludes departures that don't match ALL via stations.

Test that multiple via stations use local OR filtering and no API param.

Test that a single via station still passes when departure has no via/route fields.

Test that a single via station uses the API parameter.

Test to demonstrate potential confusion with delimiters.

Verify that different via stations result in unique fetch URLs for caching.

Test notification on cancellation.

Test that we don't notify multiple times for the same delay.

Test that a notification is triggered for a watched train.

Test that update_interval is handled correctly when set to 0.

Return the path to translations.

Handle options update.

Return UTC datetime from timestamp.

Get the walk time from config data or options.

Return the human-readable DBF website URL (without .json).