#!/usr/bin/env python3
"""
Process and simplify Dutch municipality GeoJSON for web use.

Takes the raw PDOK data and creates a simplified version suitable for web display.
Uses Shapely for proper Douglas-Peucker geometry simplification.

Input: frontend/public/data/netherlands_municipalities.geojson (raw from PDOK)
Output: frontend/public/data/netherlands_municipalities_simplified.geojson
"""

import json
from pathlib import Path
from shapely.geometry import shape, mapping
from shapely.validation import make_valid

# Paths
DATA_DIR = Path(__file__).parent.parent / "frontend" / "public" / "data"
INPUT_FILE = DATA_DIR / "netherlands_municipalities.geojson"
OUTPUT_FILE = DATA_DIR / "netherlands_municipalities_simplified.geojson"

# Simplification tolerance in degrees
# 0.001 degrees ≈ 111 meters (good for web display at typical zoom levels)
SIMPLIFY_TOLERANCE = 0.001


def simplify_feature(feature, tolerance=SIMPLIFY_TOLERANCE):
    """Simplify a GeoJSON feature using Shapely."""
    try:
        geom = shape(feature["geometry"])
        
        # Make geometry valid if needed
        if not geom.is_valid:
            geom = make_valid(geom)
        
        # Simplify using Douglas-Peucker algorithm
        simplified = geom.simplify(tolerance, preserve_topology=True)
        
        # Ensure the result is valid
        if not simplified.is_valid:
            simplified = make_valid(simplified)
        
        # Skip if geometry becomes too small/empty
        if simplified.is_empty:
            return None
        
        return {
            "type": "Feature",
            "properties": feature["properties"],
            "geometry": mapping(simplified)
        }
    except Exception as e:
        print(f"  Warning: Error simplifying {feature['properties'].get('naam', 'unknown')}: {e}")
        return feature  # Return original if simplification fails


def main():
    print(f"Loading {INPUT_FILE}...")
    with open(INPUT_FILE, "r", encoding="utf-8") as f:
        data = json.load(f)
    
    print(f"Loaded {len(data['features'])} municipalities")
    
    # Simplify each feature
    print(f"Simplifying geometries (tolerance: {SIMPLIFY_TOLERANCE} degrees ≈ {SIMPLIFY_TOLERANCE * 111000:.0f}m)...")
    simplified_features = []
    for i, feature in enumerate(data["features"]):
        if (i + 1) % 50 == 0:
            print(f"  Processed {i + 1}/{len(data['features'])}...")
        
        simplified = simplify_feature(feature)
        if simplified:
            simplified_features.append(simplified)
    
    print(f"Simplified to {len(simplified_features)} features")
    
    # Create output GeoJSON
    output = {
        "type": "FeatureCollection",
        "name": "netherlands_municipalities_simplified",
        "features": simplified_features
    }
    
    # Write with minimal whitespace for smaller file
    print(f"Writing to {OUTPUT_FILE}...")
    with open(OUTPUT_FILE, "w", encoding="utf-8") as f:
        json.dump(output, f, ensure_ascii=False, separators=(",", ":"))
    
    # Compare sizes
    input_size = INPUT_FILE.stat().st_size / (1024 * 1024)
    output_size = OUTPUT_FILE.stat().st_size / (1024 * 1024)
    reduction = (1 - output_size / input_size) * 100
    
    print(f"\nDone!")
    print(f"  Input:  {input_size:.2f} MB")
    print(f"  Output: {output_size:.2f} MB")
    print(f"  Reduction: {reduction:.1f}%")


if __name__ == "__main__":
    main()