#!/usr/bin/env python
# pylint: skip-file

import argparse
import traceback
import sys
import subprocess
import pandas as pd
import numpy as np
from sodapy import Socrata
import matplotlib.pyplot as plt


datasets_ambiente = {"2020": "nicp-bhqi", 
                     "2019": "kujm-kavy", 
                     "2018": "bgqm-yq56", 
                     "2017": "j8j8-qsb2"}


def _connect():
    client = Socrata("www.dati.lombardia.it", None)
    return client

def process_dataset(dataset, sensore):
    client = _connect()
    results = client.get(dataset, IdSensore=sensore)
    results_df = pd.DataFrame.from_records(results)
    try:
        results_df = results_df.astype({'valore': 'float64'})
    except:
        print('\nspiacente, dati non disponibili per il sensore %s') % sensore
        sys.exit(-1)
    results_df["data"] = pd.to_datetime(results_df["data"])
    results_df = results_df.replace(-9999, np.nan)
    results_df.sort_values(by=['data'], inplace=True)
    results_df.rename(columns = {'valore':sensore}, inplace = True)
    results_df.drop(columns=['idoperatore', 'idsensore', 'stato'], inplace = True)
    return results_df


def merge_df(dataframes, sensori):
    df = dataframes[sensori[0]]
    for sensore in sensori[1:]:
        df = pd.merge(df, dataframes[sensore])
    return df


def get_dataframes(datasets, sensori):
    dataframes = {}
    for sensore in sensori:
        df = process_dataset(datasets[0], sensore)
        for dataset in datasets[1:]:
            df = pd.concat([df, process_dataset(dataset, sensore)], axis=0)
        dataframes[sensore] = df
    return dataframes


def plot_dataframe(dataframe):
    dataframe.plot(x='data')
    plt.axhline(y=50, color='black', linestyle='-', label='24-hour average EU limit')
    plt.show()


def main():

    parser = argparse.ArgumentParser()
    parser.add_argument("--dataset", nargs='+', required=True, help="ricerca dei datasets")
    parser.add_argument('--sensori', nargs='+', required=True, help="specifica i sensori")
    args = parser.parse_args()

    try:
        datasets = []
        if "all" in args.dataset:
            for k in datasets_ambiente.keys():
                datasets.append(datasets_ambiente[k])
        else:
            for d in args.dataset:
                datasets.append(datasets_ambiente[d])
        sensori = args.sensori
        dataframes = get_dataframes(datasets, sensori)
        datamerged = merge_df(dataframes, sensori)
        import stazioni
        s = stazioni.get_stazioni()
        for sensore in sensori:
            location = s.loc[s['idsensore'] == sensore, 'nomestazione'].iloc[0]
            print('Valore medio per il sensore %s %s: %s' % (sensore, location, datamerged[sensore].mean().round(1)))
        plot_dataframe(datamerged)
        

    except KeyError:
        print("\nErrore:")
        #print("Datasets disponibili:")
        #print('\n'.join([str(lst) for lst in sorted(datasets_ambiente.keys())]))
        traceback.print_exc() 
    except KeyboardInterrupt:
        print("program terminated by user")
    except SystemExit:
        print("program terminated, bye")
    except:
        print("\nAn unhandled exception occured, here's the traceback!\n")
        traceback.print_exc()
        print("\nReport this to putro@autistici.org")
        sys.exit()

if __name__ == '__main__':
    main()