refactor: move FAA data conversion logic to a separate function

src/derive_from_faa_master_txt.py

@@ -0,0 +1,23 @@
+from pathlib import Path
+import zipfile
+import pandas as pd
+from faa_aircraft_registry import read
+
+def convert_faa_master_txt_to_csv(zip_path: Path, csv_path: Path) -> None:
+    with zipfile.ZipFile(zip_path) as z:
+        registrations = read(z)
+
+    df = pd.DataFrame(registrations['master'].values())
+    col = "transponder_code_hex"
+    df = df[[col] + [c for c in df.columns if c != col]]
+    df = df.rename(columns={"transponder_code_hex": "icao"})
+    registrant = pd.json_normalize(df["registrant"]).add_prefix("registrant_")
+    df = df.drop(columns="registrant").join(registrant)
+    df = df.rename(columns={"aircraft_type": "aircraft_type_2"})
+    aircraft = pd.json_normalize(df["aircraft"]).add_prefix("aircraft_")
+    df = df.drop(columns="aircraft").join(aircraft)
+    df = df.rename(columns={"engine_type": "engine_type_2"})
+    engine = pd.json_normalize(df["engine"].where(df["engine"].notna(), {})).add_prefix("engine_")
+    df = df.drop(columns="engine").join(engine)
+    df = df.sort_values(by=["icao"])
+    df.to_csv(csv_path, index=False)

src/snapshot_faa.py

@@ -1,6 +1,3 @@
-from faa_aircraft_registry import read
-import pandas as pd
-import zipfile
 import zipfile
 from pathlib import Path
 from datetime import datetime, timezone
@@ -28,21 +25,5 @@ with urlopen(req, timeout=120) as r:
     body = r.read()
     zip_path.write_bytes(body)
 
-with zipfile.ZipFile(zip_path) as z:
+from derive_from_faa_master_txt import convert_faa_master_txt_to_csv
-    registrations = read(z)
+convert_faa_master_txt_to_csv(zip_path, csv_path)
-
-df = pd.DataFrame(registrations['master'].values())
-col = "transponder_code_hex"
-df = df[[col] + [c for c in df.columns if c != col]]
-df = df.rename(columns={"transponder_code_hex": "icao"})
-registrant = pd.json_normalize(df["registrant"]).add_prefix("registrant_")
-df = df.drop(columns="registrant").join(registrant)
-df = df.rename(columns={"aircraft_type": "aircraft_type_2"})
-aircraft = pd.json_normalize(df["aircraft"]).add_prefix("aircraft_")
-df = df.drop(columns="aircraft").join(aircraft)
-df = df.rename(columns={"engine_type": "engine_type_2"})
-engine = pd.json_normalize(df["engine"].where(df["engine"].notna(), {})).add_prefix("engine_")
-df = df.drop(columns="engine").join(engine)
-df = df.sort_values(by=["icao"])
-df.to_csv(csv_path, index=False)
-