aerodatabox: add browser User-Agent, fixes 403 from Cloudflare/RapidAPI gateway

aerodatabox.py

@@ -100,8 +100,16 @@ def lookup_gate_times(flight_number: str, flight_date: date) -> dict | None:
         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}
+        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:

boarding-pass-aerodatabox-403-fix.patch

@@ -0,0 +1,44 @@
+diff --git a/aerodatabox.py b/aerodatabox.py
+index 8ba4e12..e60f4ce 100644
+--- a/aerodatabox.py
++++ b/aerodatabox.py
+@@ -100,8 +100,16 @@ def lookup_gate_times(flight_number: str, flight_date: date) -> dict | None:
+         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}
++        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:
+diff --git a/config.py b/config.py
+index 25f93bc..4056f19 100644
+--- a/config.py
++++ b/config.py
+@@ -19,10 +19,16 @@ AIRTRAIL_API_KEY = os.environ.get("AIRTRAIL_API_KEY", "")
+ # manual Search to fill in, same as today.
+ AERODATABOX_API_KEY = os.environ.get("AERODATABOX_API_KEY", "")
+ 
+-FLIGHTAWARE_USER_AGENT = (
++# Shared browser UA for every outbound scrape/API call this pipeline makes.
++# Both FlightAware and (as of the AeroDataBox 403 debugging session) the
++# AeroDataBox/RapidAPI gateway sit behind Cloudflare and reject requests
++# carrying Python urllib's default User-Agent - confirmed live: the exact
++# same request/key/URL got HTTP 403 with no UA and HTTP 200 with this one.
++BROWSER_USER_AGENT = (
+     "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 "
+     "(KHTML, like Gecko) Chrome/124.0 Safari/537.36"
+ )
++FLIGHTAWARE_USER_AGENT = BROWSER_USER_AGENT  # kept for anything still importing the old name
+ 
+ MAX_TRACK_POINTS = 2000
+ MAX_FETCH_ATTEMPTS = 5

config.py

@@ -19,10 +19,16 @@ AIRTRAIL_API_KEY = os.environ.get("AIRTRAIL_API_KEY", "")
 # manual Search to fill in, same as today.
 AERODATABOX_API_KEY = os.environ.get("AERODATABOX_API_KEY", "")
 
-FLIGHTAWARE_USER_AGENT = (
+# Shared browser UA for every outbound scrape/API call this pipeline makes.
+# Both FlightAware and (as of the AeroDataBox 403 debugging session) the
+# AeroDataBox/RapidAPI gateway sit behind Cloudflare and reject requests
+# carrying Python urllib's default User-Agent - confirmed live: the exact
+# same request/key/URL got HTTP 403 with no UA and HTTP 200 with this one.
+BROWSER_USER_AGENT = (
     "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 "
     "(KHTML, like Gecko) Chrome/124.0 Safari/537.36"
 )
+FLIGHTAWARE_USER_AGENT = BROWSER_USER_AGENT  # kept for anything still importing the old name
 
 MAX_TRACK_POINTS = 2000
 MAX_FETCH_ATTEMPTS = 5