|
| 1 | +#!/usr/bin/env python3 |
| 2 | +"""US Virgin Islands -> contributions/postcodes/VI.json importer for issue #1039. |
| 3 | +
|
| 4 | +Source data |
| 5 | +----------- |
| 6 | +USVI uses US ZIP codes in the 008xx range. The US Census ZCTA file |
| 7 | +(already shipped to contributions/postcodes/US.json under |
| 8 | +state_code='VI') contains 6 USVI-mapped postcodes with WGS-84 |
| 9 | +lat/lng centroids. |
| 10 | +
|
| 11 | +CSC represents USVI as its own country (iso2=VI, country_id=242) |
| 12 | +with 3 states (Saint Thomas / Saint John / Saint Croix). This |
| 13 | +importer mirrors the same codes into VI.json under the VI country |
| 14 | +namespace and FK'd to the nearest USVI city by centroid distance. |
| 15 | +
|
| 16 | +What this script does |
| 17 | +--------------------- |
| 18 | +1. Reads existing US.json filtered to state_code='VI' (6 codes). |
| 19 | +2. Loads contributions/cities/VI.json (20 USVI cities). |
| 20 | +3. For each VI ZIP, finds the nearest USVI city by haversine |
| 21 | + distance, uses that city's state_id (which corresponds to |
| 22 | + one of the 3 USVI islands). |
| 23 | +4. Writes contributions/postcodes/VI.json with country_id=242. |
| 24 | +
|
| 25 | +License & attribution |
| 26 | +--------------------- |
| 27 | +- Original source: US Census ZCTA Gazetteer (CC-0, public domain) |
| 28 | +- Each row: ``source: "us-census-via-vi-mirror"`` |
| 29 | +
|
| 30 | +Usage |
| 31 | +----- |
| 32 | + python3 bin/scripts/sync/import_us_virgin_islands_postcodes.py |
| 33 | +""" |
| 34 | + |
| 35 | +from __future__ import annotations |
| 36 | + |
| 37 | +import argparse |
| 38 | +import json |
| 39 | +import math |
| 40 | +import re |
| 41 | +import sys |
| 42 | +from pathlib import Path |
| 43 | +from typing import Dict, List |
| 44 | + |
| 45 | + |
| 46 | +def haversine_km(lat1: float, lon1: float, lat2: float, lon2: float) -> float: |
| 47 | + R = 6371.0 |
| 48 | + p1 = math.radians(lat1) |
| 49 | + p2 = math.radians(lat2) |
| 50 | + dlat = math.radians(lat2 - lat1) |
| 51 | + dlon = math.radians(lon2 - lon1) |
| 52 | + a = ( |
| 53 | + math.sin(dlat / 2) ** 2 |
| 54 | + + math.cos(p1) * math.cos(p2) * math.sin(dlon / 2) ** 2 |
| 55 | + ) |
| 56 | + return 2 * R * math.asin(math.sqrt(a)) |
| 57 | + |
| 58 | + |
| 59 | +def main() -> int: |
| 60 | + parser = argparse.ArgumentParser(description=__doc__) |
| 61 | + parser.add_argument("--dry-run", action="store_true") |
| 62 | + args = parser.parse_args() |
| 63 | + |
| 64 | + project_root = Path(__file__).resolve().parents[3] |
| 65 | + |
| 66 | + countries = json.load( |
| 67 | + (project_root / "contributions/countries/countries.json").open(encoding="utf-8") |
| 68 | + ) |
| 69 | + vi_country = next((c for c in countries if c.get("iso2") == "VI"), None) |
| 70 | + if vi_country is None: |
| 71 | + print("ERROR: VI not in countries.json", file=sys.stderr) |
| 72 | + return 2 |
| 73 | + regex = re.compile(vi_country.get("postal_code_regex") or ".*") |
| 74 | + |
| 75 | + us_path = project_root / "contributions/postcodes/US.json" |
| 76 | + us_data = json.load(us_path.open(encoding="utf-8")) |
| 77 | + vi_zips = [r for r in us_data if r.get("state_code") == "VI"] |
| 78 | + print(f"VI-mapped ZIPs in US.json: {len(vi_zips)}") |
| 79 | + |
| 80 | + cities_path = project_root / "contributions/cities/VI.json" |
| 81 | + vi_cities = json.load(cities_path.open(encoding="utf-8")) |
| 82 | + vi_cities_with_geo = [] |
| 83 | + for c in vi_cities: |
| 84 | + try: |
| 85 | + lat = float(c.get("latitude") or 0) |
| 86 | + lon = float(c.get("longitude") or 0) |
| 87 | + except (ValueError, TypeError): |
| 88 | + continue |
| 89 | + if lat or lon: |
| 90 | + vi_cities_with_geo.append((lat, lon, c)) |
| 91 | + print(f"VI cities with geo: {len(vi_cities_with_geo)}") |
| 92 | + |
| 93 | + states = json.load( |
| 94 | + (project_root / "contributions/states/states.json").open(encoding="utf-8") |
| 95 | + ) |
| 96 | + vi_states = {s["id"]: s for s in states if s.get("country_id") == vi_country["id"]} |
| 97 | + print( |
| 98 | + f"Country: USVI (id={vi_country['id']}); " |
| 99 | + f"states indexed: {len(vi_states)}" |
| 100 | + ) |
| 101 | + |
| 102 | + seen: set = set() |
| 103 | + records: List[dict] = [] |
| 104 | + skipped_bad_regex = 0 |
| 105 | + skipped_no_state = 0 |
| 106 | + matched_state = 0 |
| 107 | + |
| 108 | + for r in vi_zips: |
| 109 | + code = r["code"] |
| 110 | + if not regex.match(code): |
| 111 | + skipped_bad_regex += 1 |
| 112 | + continue |
| 113 | + |
| 114 | + try: |
| 115 | + lat = float(r["latitude"]) |
| 116 | + lon = float(r["longitude"]) |
| 117 | + except (ValueError, TypeError, KeyError): |
| 118 | + lat = lon = None |
| 119 | + |
| 120 | + nearest_city = None |
| 121 | + if lat is not None and lon is not None and vi_cities_with_geo: |
| 122 | + best_d = float("inf") |
| 123 | + for clat, clon, city in vi_cities_with_geo: |
| 124 | + d = haversine_km(lat, lon, clat, clon) |
| 125 | + if d < best_d: |
| 126 | + best_d = d |
| 127 | + nearest_city = city |
| 128 | + |
| 129 | + state = None |
| 130 | + locality = None |
| 131 | + if nearest_city: |
| 132 | + state = vi_states.get(nearest_city.get("state_id")) |
| 133 | + locality = nearest_city.get("name") |
| 134 | + |
| 135 | + if state is None: |
| 136 | + skipped_no_state += 1 |
| 137 | + else: |
| 138 | + matched_state += 1 |
| 139 | + |
| 140 | + key = (code, (locality or "").lower()) |
| 141 | + if key in seen: |
| 142 | + continue |
| 143 | + seen.add(key) |
| 144 | + |
| 145 | + record: Dict[str, object] = { |
| 146 | + "code": code, |
| 147 | + "country_id": int(vi_country["id"]), |
| 148 | + "country_code": "VI", |
| 149 | + } |
| 150 | + if state is not None: |
| 151 | + record["state_id"] = int(state["id"]) |
| 152 | + record["state_code"] = state.get("iso2") |
| 153 | + if locality: |
| 154 | + record["locality_name"] = locality |
| 155 | + if lat is not None and lon is not None: |
| 156 | + record["latitude"] = f"{lat:.6f}" |
| 157 | + record["longitude"] = f"{lon:.6f}" |
| 158 | + record["type"] = "full" |
| 159 | + record["source"] = "us-census-via-vi-mirror" |
| 160 | + records.append(record) |
| 161 | + |
| 162 | + print(f"Skipped (regex fail): {skipped_bad_regex:,}") |
| 163 | + print(f"Skipped (no state FK): {skipped_no_state:,}") |
| 164 | + print(f"Records emitted: {len(records):,}") |
| 165 | + pct = matched_state * 100 // max(1, len(records)) |
| 166 | + print(f" with state: {matched_state:,} ({pct}%)") |
| 167 | + |
| 168 | + if args.dry_run: |
| 169 | + return 0 |
| 170 | + |
| 171 | + target = project_root / "contributions/postcodes/VI.json" |
| 172 | + target.parent.mkdir(parents=True, exist_ok=True) |
| 173 | + if target.exists(): |
| 174 | + with target.open(encoding="utf-8") as f: |
| 175 | + existing = json.load(f) |
| 176 | + existing_seen = { |
| 177 | + (r["code"], (r.get("locality_name") or "").lower()) for r in existing |
| 178 | + } |
| 179 | + merged = list(existing) |
| 180 | + for r in records: |
| 181 | + key = (r["code"], (r.get("locality_name") or "").lower()) |
| 182 | + if key not in existing_seen: |
| 183 | + merged.append(r) |
| 184 | + existing_seen.add(key) |
| 185 | + merged.sort(key=lambda r: (r["code"], r.get("locality_name", ""))) |
| 186 | + else: |
| 187 | + merged = sorted(records, key=lambda r: (r["code"], r.get("locality_name", ""))) |
| 188 | + |
| 189 | + with target.open("w", encoding="utf-8") as f: |
| 190 | + json.dump(merged, f, ensure_ascii=False, indent=2) |
| 191 | + f.write("\n") |
| 192 | + size_kb = target.stat().st_size / 1024 |
| 193 | + print( |
| 194 | + f"\n[OK] Wrote {target.relative_to(project_root)} " |
| 195 | + f"({len(merged):,} rows, {size_kb:.0f} KB)" |
| 196 | + ) |
| 197 | + return 0 |
| 198 | + |
| 199 | + |
| 200 | +if __name__ == "__main__": |
| 201 | + raise SystemExit(main()) |
0 commit comments