fill aircraft type/registration and terminal/gate from AeroDataBox

aerodatabox.py

@@ -43,6 +43,7 @@ wheels-up/touchdown).
 import json
 import re
 import urllib.error
+import urllib.parse
 import urllib.request
 from datetime import date, datetime
 
@@ -90,6 +91,64 @@ def _utc_offset(local: str | None) -> str | None:
     return f"{offset[:3]}:{offset[3:]}" if len(offset) == 5 else None
 
 
+def _get(url: str):
+    """GET an AeroDataBox endpoint and return the decoded JSON, or None for
+    a "no content" answer. Raises AeroDataBoxError for anything retryable.
+
+    A default urllib request (no User-Agent) gets HTTP 403 from AeroDataBox's
+    Cloudflare-fronted gateway even with a perfectly valid, subscribed key -
+    confirmed live, side by side, same key/URL: bare urllib -> 403, same
+    request + this UA -> 200 with real data. See config.BROWSER_USER_AGENT."""
+    req = urllib.request.Request(
+        url,
+        headers={
+            "x-rapidapi-key": config.AERODATABOX_API_KEY,
+            "User-Agent": config.BROWSER_USER_AGENT,
+        },
+    )
+    try:
+        with urllib.request.urlopen(req, timeout=20) as resp:
+            if resp.status == 204:
+                return None
+            raw = resp.read()
+    except urllib.error.HTTPError as e:
+        if e.code == 204:
+            return None
+        raise AeroDataBoxError(f"HTTP {e.code} from AeroDataBox") from e
+    except urllib.error.URLError as e:
+        raise AeroDataBoxError(f"error reaching AeroDataBox: {e}") from e
+
+    if not raw:
+        return None
+    try:
+        return json.loads(raw.decode("utf-8"))
+    except (json.JSONDecodeError, UnicodeDecodeError) as e:
+        raise AeroDataBoxError(f"bad AeroDataBox response: {e}") from e
+
+
+def lookup_aircraft_icao(registration: str) -> str | None:
+    """Registration (e.g. 'VH-8VE') -> ICAO type code (e.g. 'B38M').
+
+    Costs one extra API unit per flight. Needed because AirTrail's
+    `aircraft` field is matched against the `icao` column of its own
+    aircraft table (getAircraftByIcao, confirmed in its source), so the
+    plain model string AeroDataBox puts on the flight record
+    ('Boeing 737 MAX 8') is not something AirTrail can resolve."""
+    if not config.AERODATABOX_API_KEY or not registration:
+        return None
+
+    data = _get(f"{BASE_URL}/aircrafts/reg/{urllib.parse.quote(registration)}")
+    if not isinstance(data, dict):
+        return None
+    # AirTrail tries model before icaoCode against the same column, so both
+    # are worth passing on in that order.
+    for key in ("icaoCode", "model"):
+        value = (data.get(key) or "").strip()
+        if value:
+            return value
+    return None
+
+
 def _first_offset(leg: dict) -> str | None:
     """The UTC offset for one leg, from whichever of its timestamps exists.
     All of a leg's times share the same airport and day, so any of them
@@ -126,35 +185,10 @@ def lookup_gate_times(flight_number: str, flight_date: date) -> dict | None:
         return None
 
     cleaned = _sanitize(flight_number)
-    url = (
+    data = _get(
         f"{BASE_URL}/flights/number/{cleaned}/{flight_date.isoformat()}"
         "?dateLocalRole=Both&withAircraftImage=false&withLocation=false"
     )
-    # A default urllib request (no User-Agent) gets HTTP 403 from AeroDataBox's
-    # Cloudflare-fronted gateway even with a perfectly valid, subscribed key -
-    # confirmed live, side by side, same key/URL: bare urllib -> 403, same
-    # request + this UA -> 200 with real data. See config.BROWSER_USER_AGENT.
-    req = urllib.request.Request(
-        url,
-        headers={
-            "x-rapidapi-key": config.AERODATABOX_API_KEY,
-            "User-Agent": config.BROWSER_USER_AGENT,
-        },
-    )
-    try:
-        with urllib.request.urlopen(req, timeout=20) as resp:
-            raw = resp.read()
-    except urllib.error.HTTPError as e:
-        if e.code == 204:
-            return None
-        raise AeroDataBoxError(f"HTTP {e.code} from AeroDataBox") from e
-    except urllib.error.URLError as e:
-        raise AeroDataBoxError(f"error reaching AeroDataBox: {e}") from e
-
-    try:
-        data = json.loads(raw.decode("utf-8"))
-    except (json.JSONDecodeError, UnicodeDecodeError) as e:
-        raise AeroDataBoxError(f"bad AeroDataBox response: {e}") from e
 
     if not isinstance(data, list) or not data:
         return None
@@ -191,4 +225,12 @@ def lookup_gate_times(flight_number: str, flight_date: date) -> dict | None:
         # into the same airport-local terms as everything else here.
         "departureUtcOffset": _first_offset(dep),
         "arrivalUtcOffset": _first_offset(arr),
+        # Straight off the flight record, no extra call. aircraftReg is a
+        # free-text field in AirTrail; the ICAO *type* code needs a separate
+        # registration lookup (see lookup_aircraft_icao).
+        "aircraftReg": ((flight.get("aircraft") or {}).get("reg") or "").strip() or None,
+        "departureTerminal": (dep.get("terminal") or "").strip() or None,
+        "departureGate": (dep.get("gate") or "").strip() or None,
+        "arrivalTerminal": (arr.get("terminal") or "").strip() or None,
+        "arrivalGate": (arr.get("gate") or "").strip() or None,
     }