test.py

import streamlit as st
import folium
import geopandas as gpd
import json
import geemap.foliumap as geemap
import ee
import pandas as pd
import os
from google.auth.transport.requests import Request
import google.auth.exceptions
from streamlit.components.v1 import html

# Path to your shapefiles and service account key
SHAPEFILE_DIR = 'C:\\Users\\piv\\Desktop\\y\\media\\shp'  # Fixed path separator

# Function to initialize GEE
def initialize_gee():
    service = os.getenv('SA')
    file = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'gee', 'ee-muzzamil1-37ebc3dece52.json')
    credentials = ee.ServiceAccountCredentials(service, file)
    try:
        ee.Initialize(credentials)
        st.success("Google Earth Engine initialized successfully.")
    except google.auth.exceptions.RefreshError:
        try:
            request = Request()
            credentials.refresh(request)
            ee.Initialize(credentials)
            st.success("Google Earth Engine token refreshed and initialized successfully.")
        except Exception as e:
            st.error(f"Error refreshing Google Earth Engine token: {e}")
    except Exception as e:
        st.error(f"Error initializing Google Earth Engine: {e}")

# Function to create Folium map with selected dataset
def create_folium_map(selected_dataset, ee_object, start_date_str, end_date_str):
    folium_map = folium.Map(location=[25.5973518, 65.54495724], zoom_start=7)

    try:
        if selected_dataset == "Modis":
            dataset = ee.ImageCollection('MODIS/006/MOD13Q1') \
                .filter(ee.Filter.date(start_date_str, end_date_str)) \
                .filterBounds(ee_object)
            def clip_image(img):
                return img.clip(ee_object).select('NDVI')
            clipped_collection = dataset.map(clip_image)
            modis_ndvi = clipped_collection.mean()
            vis_params_ndvi = {
                'min': 0,
                'max': 9000,
                'palette': ['FE8374', 'C0E5DE', '3A837C', '034B48']
            }
            modis_ndvi_map_id = modis_ndvi.getMapId(vis_params_ndvi)
            folium.TileLayer(
                tiles=modis_ndvi_map_id['tile_fetcher'].url_format,
                attr='Google Earth Engine',
                name='NDVI',
                overlay=True,
                control=True
            ).add_to(folium_map)

        elif selected_dataset == "dataset_nighttime":
            dataset_nighttime = ee.ImageCollection('NOAA/VIIRS/DNB/MONTHLY_V1/VCMCFG') \
                .filter(ee.Filter.date(start_date_str, end_date_str))
            nighttime = dataset_nighttime.select('avg_rad').mosaic()
            nighttime_clipped = nighttime.clip(ee_object)
            nighttime_vis = {
                'min': 0.0,
                'max': 60.0,
                'palette': ['1a3678', '2955bc', '5699ff', '8dbae9', 'acd1ff', 'caebff', 'e5f9ff', 'fdffb4', 'ffe6a2', 'ffc969', 'ffa12d', 'ff7c1f', 'ca531a', 'ff0000', 'ab0000']
            }
            nighttime_map_id = nighttime_clipped.getMapId(nighttime_vis)
            folium.TileLayer(
                tiles=nighttime_map_id['tile_fetcher'].url_format,
                attr='Google Earth Engine',
                name='Nighttime Lights',
                overlay=True,
                control=True
            ).add_to(folium_map)

        elif selected_dataset == "precipitation":
            dataset_precipitation = ee.ImageCollection('UCSB-CHG/CHIRPS/DAILY') \
                .filter(ee.Filter.date(start_date_str, end_date_str))
            precipitation = dataset_precipitation.mosaic().clip(ee_object)
            precip_vis = {
                'min': 0,
                'max': 300,
                'palette': ['blue', 'cyan', 'lime', 'yellow', 'red']
            }
            precip_map_id = precipitation.getMapId(precip_vis)
            folium.TileLayer(
                tiles=precip_map_id['tile_fetcher'].url_format,
                attr='Google Earth Engine',
                name='Precipitation',
                overlay=True,
                control=True
            ).add_to(folium_map)

        elif selected_dataset == "GlobalSurfaceWater":
            st.warning("GlobalSurfaceWater dataset is not configured.")
        
        elif selected_dataset == "WorldPop":
            dataset = ee.ImageCollection('WorldPop/GP/100m/pop') \
                .filterBounds(ee_object) \
                .filter(ee.Filter.date(start_date_str, end_date_str))
            population = dataset.mean().clip(ee_object)
            pop_vis = {
                'min': 0,
                'max': 1000,
                'palette': ['blue', 'green', 'yellow', 'red']
            }
            pop_map_id = population.getMapId(pop_vis)
            folium.TileLayer(
                tiles=pop_map_id['tile_fetcher'].url_format,
                attr='Google Earth Engine',
                name='World Population',
                overlay=True,
                control=True
            ).add_to(folium_map)

        elif selected_dataset == "COPERNICUS":
            dataset = ee.ImageCollection('COPERNICUS/S2') \
                .filterBounds(ee_object) \
                .filter(ee.Filter.date(start_date_str, end_date_str))
            sentinel = dataset.median().clip(ee_object)
            vis_params = {
                'bands': ['B4', 'B3', 'B2'],  # RGB bands for visualization
                'min': 0,
                'max': 3000,
                'gamma': 1.4
            }
            copernicus_map_id = sentinel.getMapId(vis_params)
            folium.TileLayer(
                tiles=copernicus_map_id['tile_fetcher'].url_format,
                attr='Google Earth Engine',
                name='Copernicus Sentinel-2',
                overlay=True,
                control=True
            ).add_to(folium_map)

        folium_map.add_child(folium.LayerControl())
    except Exception as e:
        st.error(f"Error processing dataset: {e}")
        st.write(f"Exception: {e}")

    # Convert Folium map to HTML
    folium_map_html = folium_map._repr_html_()
    st.write(folium_map_html)  # Debug line
    return folium_map_html

# Function to handle submit button logic
def handle_submit(selected_dataset, selected_date_range_From, selected_date_range_To, selected_shape):
    # Construct shapefile path based on selected shape
    shapefile_name = f'{selected_shape}.shp'
    shapefile_path = os.path.join(SHAPEFILE_DIR, shapefile_name)

    if os.path.exists(shapefile_path):
        try:
            # Read shapefile
            roi_gdf = gpd.read_file(shapefile_path)
            roi_geojson = roi_gdf.to_crs("EPSG:4326").to_json()
        except Exception as e:
            st.error(f"Error reading shapefile: {e}")
            return
        
        # Convert GeoJSON to Earth Engine object
        try:
            ee_object = geemap.geojson_to_ee(json.loads(roi_geojson))
        except Exception as e:
            st.error(f"Error converting GeoJSON to Earth Engine object: {e}")
            return
        
        # Create and display Folium map for satellite data
        st.markdown("### Satellite Data Map")
        folium_map_html = create_folium_map(
            selected_dataset,
            ee_object,
            selected_date_range_From.strftime('%Y-%m-%d'),
            selected_date_range_To.strftime('%Y-%m-%d')
        )
        # Use Streamlit to render the HTML
        html(folium_map_html, height=600)
    else:
        st.error("Selected shapefile does not exist.")

# Streamlit application
def main():
    st.set_page_config(layout="wide")
    st.title("Satellite Data Visualization, Interactive Map")

    # Initialize GEE when the app starts
    initialize_gee()

    # Sidebar with general information
    st.sidebar.title("Controls")

    submit_control()   

    # Dataset selection

def submit_control():   
    selected_dataset = st.sidebar.selectbox(
    "Select Dataset", 
    ["Modis", "dataset_nighttime", 'precipitation', 'GlobalSurfaceWater', 'WorldPop', 'COPERNICUS']
    )

    # Date range selection
    selected_date_range_From = st.sidebar.date_input("From", value=pd.to_datetime("2015-07-01"))
    selected_date_range_To = st.sidebar.date_input("To", value=pd.to_datetime("2023-09-30"))

    # Shape options
    shape_options = [
    'District_Boundary', 'hydro_basins', 'karachi', 
    'National_Constituency_with_Projected_2010_Population', 
    'Provincial_Boundary', 'Provincial_Constituency', 
    'Tehsil_Boundary', 'Union_Council'
    ]

    # Shape selection
    selected_shape = st.sidebar.selectbox("Select Shape", shape_options)

    # Add a submit button
    if st.sidebar.button("Submit"):
        # Store the current inputs in session state
        st.session_state.selected_dataset = selected_dataset
        st.session_state.selected_date_range_From = selected_date_range_From
        st.session_state.selected_date_range_To = selected_date_range_To
        st.session_state.selected_shape = selected_shape
     
    # Check if the form has been submitted
    if 'selected_dataset' in st.session_state:
        handle_submit(
            st.session_state.selected_dataset,
            st.session_state.selected_date_range_From,
            st.session_state.selected_date_range_To,
            st.session_state.selected_shape
        )
     

if __name__ == "__main__":
    main()