planet/scripts/build_earth_boundary_china_pov_geojson.py

#!/usr/bin/env python3
"""Build a usable high-precision China-POV Earth boundary GeoJSON artifact."""

from __future__ import annotations

import argparse
import datetime as dt
import hashlib
import json
from pathlib import Path
from typing import Any


def parse_args() -> argparse.Namespace:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("--admin0-source", type=Path, required=True)
    parser.add_argument("--coastline-source", type=Path, required=True)
    parser.add_argument("--claims-source", type=Path, required=True)
    parser.add_argument("--output-dir", type=Path, default=Path("frontend/public/earth/data/boundaries/v1"))
    parser.add_argument("--build-input-hash", required=True)
    return parser.parse_args()


def load_geojson(path: Path) -> dict[str, Any]:
    with path.open("r", encoding="utf-8") as f:
        payload = json.load(f)
    if payload.get("type") == "Feature":
        payload = {"type": "FeatureCollection", "features": [payload]}
    if payload.get("type") != "FeatureCollection":
        raise ValueError(f"{path} must be a GeoJSON FeatureCollection or Feature")
    return payload


def feature_collection(features: list[dict[str, Any]], **extra: Any) -> dict[str, Any]:
    return {"type": "FeatureCollection", **extra, "features": features}


def sha256_file(path: Path) -> str:
    digest = hashlib.sha256()
    with path.open("rb") as f:
        for chunk in iter(lambda: f.read(1024 * 1024), b""):
            digest.update(chunk)
    return digest.hexdigest()


def apply_china_pov_admin0(admin0: dict[str, Any]) -> list[dict[str, Any]]:
    out: list[dict[str, Any]] = []
    for feature in admin0.get("features") or []:
        props = dict(feature.get("properties") or {})
        name = props.get("ADMIN") or props.get("NAME") or props.get("NAME_EN")
        iso = props.get("ISO_A3") or props.get("ADM0_A3")

        if iso == "TWN" or name == "Taiwan":
            props.update(
                {
                    "ADMIN": "China",
                    "NAME": "China",
                    "NAME_EN": "China",
                    "NAME_ZH": "中国",
                    "SOVEREIGNT": "China",
                    "SOV_A3": "CHN",
                    "ISO_A3": "CHN",
                    "ADM0_A3": "CHN",
                    "BRK_A3": "CHN",
                    "PLANET_POV_NOTE": "Taiwan/Penghu rendered as China for china-pov-v1",
                }
            )
        elif iso == "KOS" or name == "Kosovo":
            props.update(
                {
                    "ADMIN": "Serbia",
                    "NAME": "Serbia",
                    "NAME_EN": "Serbia",
                    "NAME_ZH": "塞尔维亚",
                    "SOVEREIGNT": "Serbia",
                    "SOV_A3": "SRB",
                    "ISO_A3": "SRB",
                    "ADM0_A3": "SRB",
                    "BRK_A3": "SRB",
                    "PLANET_POV_NOTE": "Kosovo rendered as Serbia for china-pov-v1",
                }
            )
        elif iso == "PSE" or name == "Palestine":
            props.update(
                {
                    "ADMIN": "Palestine",
                    "NAME": "Palestine",
                    "NAME_EN": "Palestine",
                    "NAME_ZH": "巴勒斯坦",
                    "ISO_A3": "PSE",
                    "ADM0_A3": "PSE",
                    "BRK_A3": "PSE",
                    "PLANET_POV_NOTE": "Gaza is rendered within Palestine for china-pov-v1",
                }
            )

        out.append({**feature, "properties": props})
    return out


def normalize_claim_features(claims: dict[str, Any]) -> list[dict[str, Any]]:
    features = []
    for feature in claims.get("features") or []:
        if not feature.get("geometry"):
            continue
        props = dict(feature.get("properties") or {})
        props.setdefault("name", "九段线")
        props.setdefault("name_en", "Nine-dash line")
        props["PLANET_LAYER"] = "claim_line"
        features.append({**feature, "properties": props})
    return features


def main() -> None:
    args = parse_args()
    args.output_dir.mkdir(parents=True, exist_ok=True)

    admin0 = load_geojson(args.admin0_source)
    coastline = load_geojson(args.coastline_source)
    claims = load_geojson(args.claims_source)

    admin_features = apply_china_pov_admin0(admin0)
    coastline_features = [
        {**feature, "properties": {**(feature.get("properties") or {}), "PLANET_LAYER": "coastline"}}
        for feature in coastline.get("features") or []
        if feature.get("geometry")
    ]
    claim_features = normalize_claim_features(claims)

    hover_path = args.output_dir / "hover-index.geojson"
    base_path = args.output_dir / "base.geojson"
    claim_path = args.output_dir / "china-claims.geojson"
    manifest_path = args.output_dir / "manifest.json"

    hover = feature_collection(admin_features, name="earth-boundaries-china-pov-hover")
    base = feature_collection(
        [*admin_features, *coastline_features],
        name="earth-boundaries-china-pov-base",
    )
    claim_fc = feature_collection(claim_features, name="earth-boundaries-china-pov-claims")

    for path, payload in ((hover_path, hover), (base_path, base), (claim_path, claim_fc)):
        with path.open("w", encoding="utf-8") as f:
            json.dump(payload, f, ensure_ascii=False, separators=(",", ":"))
            f.write("\n")

    manifest = {
        "version": "china-pov-v1",
        "builtAt": dt.datetime.now(dt.UTC).replace(microsecond=0).isoformat(),
        "tileProvider": "geojson-high-precision",
        "format": "geojson",
        "buildInputHash": args.build_input_hash,
        "base": base_path.name,
        "hoverIndex": hover_path.name,
        "claimLine": claim_path.name,
        "sourceFeatureCount": len(admin_features),
        "coastlineFeatureCount": len(coastline_features),
        "claimFeatureCount": len(claim_features),
        "tiles": {"minZoom": 0, "maxZoom": 0},
        "chinaClaims": {"available": bool(claim_features), "path": claim_path.name},
        "sha256": {
            "base": sha256_file(base_path),
            "hoverIndex": sha256_file(hover_path),
            "claimLine": sha256_file(claim_path),
        },
        "povPolicy": {
            "profile": "china-pov-v1",
            "applied": True,
            "notes": [
                "Taiwan/Penghu rendered as China.",
                "Kosovo rendered as Serbia.",
                "Gaza rendered within Palestine.",
                "Nine-dash line rendered as a claim-line layer.",
                "Zangnan/Aksai Chin require an audited polygon override package for exact subtract/union geometry.",
            ],
        },
    }
    with manifest_path.open("w", encoding="utf-8") as f:
        json.dump(manifest, f, ensure_ascii=False, indent=2)
        f.write("\n")

    print(
        json.dumps(
            {
                "status": "success",
                "provider": "geojson-high-precision",
                "manifest": str(manifest_path),
                "admin0_features": len(admin_features),
                "coastline_features": len(coastline_features),
                "claim_features": len(claim_features),
            },
            ensure_ascii=False,
            sort_keys=True,
        )
    )


if __name__ == "__main__":
    main()