|
| 1 | +#!/usr/bin/env python3 |
| 2 | +"""Australia Post -> contributions/postcodes/AU.json importer for issue #1039. |
| 3 | +
|
| 4 | +Source data |
| 5 | +----------- |
| 6 | +The community-maintained Matthew Proctor archive is the standard |
| 7 | +redistributable mirror of Australia Post's open-data postcode list, |
| 8 | +published under CC-BY 4.0 attribution to Australia Post: |
| 9 | +
|
| 10 | + https://github.com/matthewproctor/australianpostcodes |
| 11 | +
|
| 12 | +The CSV has ~18,500 rows representing localities; ~3,175 unique postcodes |
| 13 | +spread across the 8 Australian states/territories. |
| 14 | +
|
| 15 | +What this script does |
| 16 | +--------------------- |
| 17 | +1. Reads the CSV (UTF-8, header row) |
| 18 | +2. Picks ONE canonical record per unique postcode (first alphabetical |
| 19 | + locality name — gives a stable primary that future curated PRs can |
| 20 | + override per-row) |
| 21 | +3. Resolves country_id (AU) and state_id by direct iso2 match |
| 22 | + (CSV uses ACT/NSW/NT/QLD/SA/TAS/VIC/WA — same codes as states.json) |
| 23 | +4. Carries forward latitude/longitude when present |
| 24 | +5. Writes contributions/postcodes/AU.json |
| 25 | +6. Idempotent merge with existing curated rows |
| 26 | +
|
| 27 | +License & attribution |
| 28 | +--------------------- |
| 29 | +- Upstream source: Australia Post (CC-BY 4.0) |
| 30 | +- Mirror: github.com/matthewproctor/australianpostcodes |
| 31 | +- Each generated row records source: "australia-post" |
| 32 | +
|
| 33 | +Usage |
| 34 | +----- |
| 35 | + curl -L -o /tmp/au_postcodes.csv \\ |
| 36 | + https://raw.githubusercontent.com/matthewproctor/australianpostcodes/master/australian_postcodes.csv |
| 37 | +
|
| 38 | + python3 bin/scripts/sync/import_australia_post_postcodes.py |
| 39 | +""" |
| 40 | + |
| 41 | +from __future__ import annotations |
| 42 | + |
| 43 | +import argparse |
| 44 | +import csv |
| 45 | +import json |
| 46 | +import sys |
| 47 | +from pathlib import Path |
| 48 | +from typing import Dict, List, Optional |
| 49 | + |
| 50 | + |
| 51 | +def parse_coord(v: str) -> Optional[str]: |
| 52 | + if not v or not v.strip(): |
| 53 | + return None |
| 54 | + try: |
| 55 | + f = float(v) |
| 56 | + if abs(f) > 180: |
| 57 | + return None |
| 58 | + return f"{f:.8f}".rstrip("0").rstrip(".") or "0" |
| 59 | + except ValueError: |
| 60 | + return None |
| 61 | + |
| 62 | + |
| 63 | +def main() -> int: |
| 64 | + parser = argparse.ArgumentParser(description=__doc__) |
| 65 | + parser.add_argument("--input", default="/tmp/au_postcodes.csv") |
| 66 | + parser.add_argument("--dry-run", action="store_true") |
| 67 | + args = parser.parse_args() |
| 68 | + |
| 69 | + csv_path = Path(args.input) |
| 70 | + if not csv_path.exists(): |
| 71 | + print(f"ERROR: input not found: {csv_path}", file=sys.stderr) |
| 72 | + return 2 |
| 73 | + |
| 74 | + project_root = Path(__file__).resolve().parents[3] |
| 75 | + countries = json.load((project_root / "contributions/countries/countries.json").open(encoding="utf-8")) |
| 76 | + au = next((c for c in countries if c.get("iso2") == "AU"), None) |
| 77 | + if au is None: |
| 78 | + print("ERROR: AU not in countries.json", file=sys.stderr) |
| 79 | + return 2 |
| 80 | + states = json.load((project_root / "contributions/states/states.json").open(encoding="utf-8")) |
| 81 | + au_states = [s for s in states if s.get("country_id") == au["id"]] |
| 82 | + state_by_iso2: Dict[str, dict] = {(s.get("iso2") or "").upper(): s for s in au_states if s.get("iso2")} |
| 83 | + |
| 84 | + print(f"Country: Australia (id={au['id']}); states: {len(au_states)}") |
| 85 | + |
| 86 | + # First pass: group rows by postcode, then pick a canonical one per postcode |
| 87 | + by_postcode: Dict[str, List[dict]] = {} |
| 88 | + with csv_path.open(encoding="utf-8") as f: |
| 89 | + reader = csv.DictReader(f) |
| 90 | + for row in reader: |
| 91 | + code = (row.get("postcode") or "").strip() |
| 92 | + if not code or not code.isdigit() or len(code) != 4: |
| 93 | + continue |
| 94 | + by_postcode.setdefault(code, []).append(row) |
| 95 | + |
| 96 | + # Sort each postcode's rows by locality name; first wins |
| 97 | + records: List[dict] = [] |
| 98 | + matched_state = 0 |
| 99 | + matched_coord = 0 |
| 100 | + for code in sorted(by_postcode): |
| 101 | + rows = sorted(by_postcode[code], key=lambda r: (r.get("locality") or "").upper()) |
| 102 | + chosen = rows[0] |
| 103 | + record = { |
| 104 | + "code": code, |
| 105 | + "country_id": int(au["id"]), |
| 106 | + "country_code": "AU", |
| 107 | + } |
| 108 | + st_iso = (chosen.get("state") or "").strip().upper() |
| 109 | + state = state_by_iso2.get(st_iso) |
| 110 | + if state is not None: |
| 111 | + record["state_id"] = int(state["id"]) |
| 112 | + record["state_code"] = st_iso |
| 113 | + matched_state += 1 |
| 114 | + locality = (chosen.get("locality") or "").strip() |
| 115 | + if locality: |
| 116 | + record["locality_name"] = locality |
| 117 | + record["type"] = "full" |
| 118 | + lat = parse_coord(chosen.get("lat") or "") |
| 119 | + lng = parse_coord(chosen.get("long") or "") |
| 120 | + if lat is not None and lng is not None: |
| 121 | + record["latitude"] = lat |
| 122 | + record["longitude"] = lng |
| 123 | + matched_coord += 1 |
| 124 | + record["source"] = "australia-post" |
| 125 | + records.append(record) |
| 126 | + |
| 127 | + print(f"Records: {len(records):,}") |
| 128 | + print(f" with state: {matched_state:,} ({matched_state*100//max(1,len(records))}%)") |
| 129 | + print(f" with coords: {matched_coord:,} ({matched_coord*100//max(1,len(records))}%)") |
| 130 | + |
| 131 | + if args.dry_run: |
| 132 | + return 0 |
| 133 | + |
| 134 | + target = project_root / "contributions/postcodes/AU.json" |
| 135 | + if target.exists(): |
| 136 | + with target.open(encoding="utf-8") as f: |
| 137 | + existing = json.load(f) |
| 138 | + seen = {(r["code"], (r.get("locality_name") or "").lower()) for r in existing} |
| 139 | + merged = list(existing) |
| 140 | + for r in records: |
| 141 | + key = (r["code"], (r.get("locality_name") or "").lower()) |
| 142 | + if key not in seen: |
| 143 | + merged.append(r) |
| 144 | + seen.add(key) |
| 145 | + merged.sort(key=lambda r: (r["code"], r.get("locality_name", ""))) |
| 146 | + else: |
| 147 | + merged = sorted(records, key=lambda r: (r["code"], r.get("locality_name", ""))) |
| 148 | + |
| 149 | + with target.open("w", encoding="utf-8") as f: |
| 150 | + json.dump(merged, f, ensure_ascii=False, indent=2) |
| 151 | + f.write("\n") |
| 152 | + size_kb = target.stat().st_size / 1024 |
| 153 | + print(f"\n[OK] Wrote {target.relative_to(project_root)} ({len(merged):,} rows, {size_kb:.0f} KB)") |
| 154 | + return 0 |
| 155 | + |
| 156 | + |
| 157 | +if __name__ == "__main__": |
| 158 | + raise SystemExit(main()) |
0 commit comments