import pandas as pd
import numpy as np
from pathlib import Path
import json


base_dir = Path(__file__).parent.parent
filename = base_dir / 'data' / 'Marketingplanung_AHR_2023_V1.xls'
output = base_dir / 'export' / 'marketing_2023.json'
output_csv = base_dir / 'export' / 'marketing_2023.csv'
debug_csv = base_dir / 'export' / 'marketing_2023_rest.csv'

department = {
    ' MM': '10', 
    ' ULM': '40', 
    ' LL': '50', 
    'KRU': '30', 
    'GZ': '55', 
    'AAM': '82', 
    'AAM-MOT': '81'
}

columns = [
    'Bezeichnung', 'Segment', 'Termin', 
    'Wert1', 'Wert2', 'Wert3', 'Wert4', 'Wert5', 'Wert6',
    'Wert7', 'Wert8', 'Wert9', 'Wert10', 'Wert11', 'Wert12', 'Summe'
]

defaults = ['', '', '', 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]
types = [str] * 3  # + [float] * 13

xls = pd.read_excel(
    filename, 
    sheet_name=list(department.keys()),
    skiprows=4,
    usecols='A:P',
    names=columns,
    converters=dict(zip(columns, types)),
)
xls_temp = []

for sheet, dept_no in department.items():
    xls[sheet]['Betrieb Nr'] = dept_no
    xls[sheet] = xls[sheet].fillna(value=dict(zip(columns, defaults)))
    xls[sheet]['Konto Nr'] = np.where(xls[sheet]['Segment'].str.startswith('Kto '), xls[sheet]['Segment'].str.slice(4, 9), 'Rest')
    # xls[sheet]['Konto Nr'] = np.where(xls[sheet]['Segment'].str.startswith('50% '), xls[sheet]['Segment'].str.slice(-5), xls[sheet]['Konto Nr'])
    temp = xls[sheet][xls[sheet]['Segment'].str.startswith('50% ')].copy()
    temp['Konto Nr'] = temp['Segment'].str.slice(8, 13)
    xls_temp.append(temp)
    temp2 = temp.copy()
    temp2['Konto Nr'] = temp2['Segment'].str.slice(-5)
    xls_temp.append(temp2)

xls_temp_concat = pd.concat(xls_temp)
for c in columns[3:]:
    xls_temp_concat[c] = xls_temp_concat[c] * 0.5

df = pd.concat(list(xls.values()) + [xls_temp_concat])

df_rest = df[(df['Konto Nr'] == 'Rest') & (df['Summe'] != 0)]
df_rest.to_csv(debug_csv, sep=';', decimal=',')
df = df[df['Konto Nr'] != 'Rest']

df['Wert11'] = df['Wert11'].astype(float)
df['Wert12'] = df['Wert12'].astype(float)

for c in columns[3:]:
    df[c] = df[c].round()
df = df.drop(labels=['Summe'], axis=1)

df = df.groupby(['Konto Nr', 'Betrieb Nr']).sum()
print(df.info())
# df.reset_index(inplace=True)

df.to_csv(output_csv, sep=';', decimal=',')
dict_split = df.to_dict(orient='split')

res = {}
for index, data in zip(dict_split['index'], dict_split['data']):
    kto, dept = index
    if kto not in res:
        res[kto] = {}
    res[kto][dept] = data

json.dump(res, open(output, 'w'), indent=2)