Federate official Belgium data sources

scripts/provision_mol_population_history.py

@@ -255,6 +255,23 @@ def resolve_boundary_path(args: argparse.Namespace, scope: GeographicScope) -> P
         return Path(legacy)
     if scope.key == "mol" and DEFAULT_BOUNDARY_PATH.exists():
         return DEFAULT_BOUNDARY_PATH
+    if scope.key == "belgium":
+        national_scope_dir = args.scope_output_root / "belgium-north-sea"
+        manifest_path = national_scope_dir / "manifest.json"
+        boundary_path = national_scope_dir / "belgium_land_boundary.geojson"
+        if not manifest_path.is_file() or not boundary_path.is_file():
+            raise RuntimeError(
+                "Official Belgium boundary evidence is missing; "
+                "run provision_belgium_north_sea_scope.py --fetch-only first"
+            )
+        manifest = json.loads(manifest_path.read_text(encoding="utf-8"))
+        artifact = (manifest.get("artifacts") or {}).get("belgium_land_boundary") or {}
+        if (
+            manifest.get("scope") != "belgium-and-belgian-north-sea"
+            or artifact.get("sha256") != sha256_path(boundary_path)
+        ):
+            raise RuntimeError("Official Belgium boundary evidence does not match the retained manifest")
+        return boundary_path
     scope_dir = args.scope_output_root / scope.key
     manifest_path = scope_dir / f"{scope.key.replace('-', '_')}_scope_manifest.json"
     if not manifest_path.exists():
@@ -330,13 +347,14 @@ def scoped_population_rows(source_rows: list[dict[str, Any]], scope: GeographicS
         if sector_code in seen:
             raise RuntimeError(f"Statbel population table contains duplicate sector {sector_code}")
         seen.add(sector_code)
-        if nis_code not in members:
+        if not scope.all_municipalities and nis_code not in members:
             continue
+        municipality_name = members.get(nis_code) or str(row.get("TX_DESCR_NL") or nis_code).strip()
         rows[sector_code] = {
             "population_total": int(total_raw),
             "sector_name_nl": row.get("TX_DESCR_SECTOR_NL"),
             "municipality_name_nl": row.get("TX_DESCR_NL"),
-            "municipality": members[nis_code],
+            "municipality": municipality_name,
             "nis_code": nis_code,
         }
     if not rows:
@@ -358,7 +376,10 @@ def build_snapshot(
     missing_population = 0
     for source_feature in sector_payload.get("features") or []:
         properties = source_feature.get("properties") or {}
-        if str(properties.get("cd_munty_refnis") or "") not in member_codes:
+        if (
+            not scope.all_municipalities
+            and str(properties.get("cd_munty_refnis") or "") not in member_codes
+        ):
             continue
         sector_code = str(properties.get("cd_sector") or "").strip()
         population_values = population.get(sector_code)
@@ -406,8 +427,8 @@ def build_snapshot(
         "features": features,
         "coverage_scope": scope.key,
         "scope_type": scope.scope_type,
-        "member_count": len(scope.members),
-        "member_nis_codes": list(scope.nis_codes),
+        "member_count": int(accounting.get("member_count") or len(scope.members)),
+        "member_nis_codes": list(accounting.get("member_nis_codes") or scope.nis_codes),
         "geometry_clipped_to_area": True,
         "observation_year": year,
         "missing_population_sector_count": missing_population,
@@ -594,14 +615,21 @@ def upload_snapshot(
     observed_at = f"{year}-01-01T00:00:00Z"
     preflight = load_preflight_manifest(preflight_path, path, year, scope)
     accounting = preflight["scope_accounting"]
+    coverage_zones = (
+        ["belgium", "flanders", "wallonia", "brussels"]
+        if scope.key == "belgium"
+        else ["flanders"]
+    )
     source_metadata = {
         "provider": "Statbel",
         "authority_level": "authoritative",
+        "theme": "population",
+        "coverage_zones": coverage_zones,
         "coverage_scope": scope.key,
         "scope_type": scope.scope_type,
         "scope_display_name": scope.display_name,
-        "member_count": len(scope.members),
-        "member_nis_codes": list(scope.nis_codes),
+        "member_count": int(accounting.get("member_count") or len(scope.members)),
+        "member_nis_codes": list(accounting.get("member_nis_codes") or scope.nis_codes),
         "geometry_clipped_to_area": True,
         "attribution": ATTRIBUTION,
         "license": "CC BY 4.0",
@@ -742,6 +770,7 @@ def main() -> int:
                         "path": path,
                         "manifest_path": manifest_path if manifest_path.is_file() else None,
                         "feature_count": len(payload.get("features") or []),
+                        "member_count": int(payload.get("member_count") or len(scope.members)),
                         "preflight_status": preflight_status,
                     }
                 )
@@ -814,7 +843,10 @@ def main() -> int:
                 "status": "ok",
                 "scope": scope.key,
                 "display_name": scope.display_name,
-                "member_count": len(scope.members),
+                "member_count": max(
+                    (int(item.get("member_count") or 0) for item in prepared),
+                    default=len(scope.members),
+                ),
                 "series": series_key(scope),
                 "snapshots": results,
             },