import numpy as np
import cv2
import json
from sentinelhub import SHConfig, MimeType, BBox, SentinelHubRequest, SentinelHubDownloadClient, DataCollection, geo_utils

def detect_deforestation(bbox, time_interval, output_dir, config=None):
    if config is None:
        # Configure SentinelHub with your credentials
        config = SHConfig()
        config.sh_client_id = '3e9277b3-aa11-4f98-a160-3bf006130cc7'
        config.sh_client_secret = 'CY>imU7J,97EniGj5jyRtz:l13@XL2*gfY+pvqRw'
    
    # Define the SentinelHub request parameters
    evalscript = """
        // Import Sentinel-2 bands
        return [B02, B03, B04, B08];
    """

    request = SentinelHubRequest(
        evalscript=evalscript,
        input_data=[
            SentinelHubRequest.input_data(
                data_collection=DataCollection.SENTINEL2_L1C,
                time_interval=time_interval,
                maxcc=0.2
            )
        ],
        responses=[
            SentinelHubRequest.output_response('default', MimeType.TIFF),
        ],
        bbox=bbox,
        config=config,
    )

    # Fetch Sentinel imagery
    sentinel_data = request.get_data()
    sentinel_data = np.array(sentinel_data[0])

    # Calculate the NDVI (Normalized Difference Vegetation Index)
    nir = sentinel_data[:, :, 3]
    red = sentinel_data[:, :, 2]
    ndvi = (nir - red) / (nir + red)

    # Apply a threshold to identify deforested areas
    threshold = 0.2
    deforestation_mask = (ndvi < threshold)

    # Create a color-coded PNG image
    deforestation_image = np.zeros_like(sentinel_data)
    deforestation_image[deforestation_mask] = [0, 0, 255, 255]  # Red color for deforestation

    # Save the PNG image
    png_path = f"{output_dir}/deforestation.png"
    cv2.imwrite(png_path, deforestation_image)

    # Create a GeoJSON object to represent deforested areas
    # geojson_features = []
    # rows, cols = deforestation_mask.shape
    # for row in range(rows):
    #     for col in range(cols):
    #         if deforestation_mask[row, col]:
    #             # Convert pixel coordinates to geographical coordinates
    #             #lon, lat = request.get_bbox().pixel_to_utm((row, col))
    #             #lon, lat = geo_utils.epsg_to_wgs84((bbox.min_x + col * request.resolution_x, bbox.min_y + row * request.resolution_y))
    #             lon, lat = bbox.pixel_to_utm((col, row))
    #             geojson_features.append({
    #                 "type": "Feature",
    #                 "geometry": {
    #                     "type": "Point",
    #                     "coordinates": [lon, lat]
    #                 }
    #             })

    # geojson = {
    #     "type": "FeatureCollection",
    #     "features": geojson_features
    # }

    # # Save the GeoJSON object
    # geojson_path = f"{output_dir}/deforestation.geojson"
    # with open(geojson_path, "w") as geojson_file:
    #     json.dump(geojson, geojson_file)

    return png_path#, geojson_path

# Example usage:

longitude_min, latitude_min, longitude_max, latitude_max = 103.88372272,-2.34602829,103.94000217,-2.41710372
bbox = BBox([longitude_min, latitude_min, longitude_max, latitude_max], crs="EPSG:4326")
time_interval = ('2022-01-01', '2023-07-31')
output_directory = 'SFTP'
result = detect_deforestation(bbox, time_interval, output_directory)
print(result)