import os
from pathlib import Path

from gisco_geodata import (
    NUTS,
    Countries,
    set_httpx_args,
    set_semaphore_value
)


if __name__ == '__main__':
    out_dir = Path(
        os.path.normpath(os.path.expanduser("~/Desktop"))
    )  # Desktop path

    set_httpx_args(verify=False)  # prevents SSLError in my case
    set_semaphore_value(20) # the maximum number of asynchronous API calls

    nuts = NUTS()

    nuts.download(
        file_format='shp',
        year='2021',
        spatial_type='BN',
        scale='60M',
        projection='4326',
        nuts_level='LEVL_3',
        out_dir=out_dir,
    )

    # Equivalent to the above
    datasets = nuts.get_datasets()
    datasets[-1].download(
        file_format='shp',
        spatial_type='BN',
        scale='60M',
        projection='4326',
        nuts_level='LEVL_3',
        out_dir=out_dir,
    )

    # Retrieve Country information as Polygons
    countries = Countries()

    # If you have geopandas installed, this will be a GDF.
    gdf = countries.get(
        countries=['RO', 'IT'],
        spatial_type='RG',
    )
    if not isinstance(gdf, list):
        print(gdf.head(5))
    else:
        print(gdf)

"""To use this script you will need to also install the 'eurostat' and the 'mapclassify' packages."""

from gisco_geodata import (
    NUTS,
    set_httpx_args
)
from eurostat import (
    get_data_df,
    get_toc_df,
    set_requests_args
)
import geopandas as gpd
import matplotlib.pyplot as plt


if __name__ == "__main__":
    set_httpx_args(verify=False)
    set_requests_args(verify=False)

    # Get the geometries from the gisco service.
    nuts = NUTS()
    level_2 = nuts.get(spatial_type='RG', nuts_level='LEVL_2')
    assert isinstance(level_2, gpd.GeoDataFrame)

    # Get the dataset information.
    eurostat_database = get_toc_df()
    code = eurostat_database.loc[eurostat_database['title'] == 'Unemployment rate by NUTS 2 regions', 'code'].iloc[0]
    dataset = get_data_df(code)
    assert dataset is not None

    # Preprocess the dataset.
    dataset = dataset.loc[(dataset['isced11'] == 'TOTAL') & (dataset['sex'] == 'T')]  # total unemployment rate

    # Join with the geometries.
    dataset = level_2.merge(dataset, left_on='FID', right_on=r'geo\TIME_PERIOD')
    assert isinstance(dataset, gpd.GeoDataFrame)

    # Plot.
    dataset.plot(column='2023', scheme='NaturalBreaks', legend=True, edgecolor='black',
                 title='Unemployment rate by NUTS 2 regions, 2023')
    plt.show()