# All dataframe calculations

import pandas as pd

MAX_DECIMAL_POINTS=10

# Generate empty dataframe with final report structure
def createEmptyDf() -> pd.DataFrame:
    return pd.DataFrame({
        'year_month': pd.Series([], dtype='str'),
        'month_cost': pd.Series([], dtype='float'),
        'environment': pd.Series([], dtype='str'),
        'resources': pd.Series([], dtype='str'),
        'daily_cost': pd.Series([], dtype='float'),
        'cost': pd.Series([], dtype='float'),
        'date': pd.Series([], dtype='str'),
        'product': pd.Series([], dtype='str'),
        'daily_product_cost': pd.Series([], dtype='float'),
        'daily_environment_cost': pd.Series([], dtype='float'),
    })


# From given untagged service dataframe generate dataframe with structure of final output
# all summary fields set to 0
def generateAccountCost(in_df: pd.DataFrame, account: str, env: str) -> pd.DataFrame:
    out_df = createEmptyDf()

    in_dict = in_df.to_dict()
    for service, row in in_dict.items():
        service = service.replace('($)','').lower()
        resources = 'untagged_' + service.replace(' ','_').lower()

        next_frame = pd.DataFrame.from_dict({
            'year_month': pd.Series([date.rsplit('-', 1)[0] for date in row.keys()], dtype='str'),
            'month_cost': 0,
            'environment': env,
            'resources': resources,
            'daily_cost': 0,
            'cost': pd.Series([round(x, MAX_DECIMAL_POINTS) for x in row.values()], dtype='float'),
            'date': pd.Series(row.keys(), dtype='str'),
            'product': account,
            'daily_product_cost': 0,
            'daily_environment_cost': 0,
        },)
        out_df = pd.concat([out_df, next_frame], ignore_index=True)
    return out_df

# From given tagged dataframe generate dataframe with structure of final output
# all summary fields set to 0
def generateAmortizedCosts(in_df: pd.DataFrame) -> pd.DataFrame:
    out_df = createEmptyDf()
    in_dict = in_df.to_dict()
    for tag, row in in_dict.items():
        tag = tag.replace('($)','').lower()
        # one broken tag fix
        tag = tag.replace('-','_')
        # extend names for cmn env and apl and dlp products
        splited_tag = tag.split('_', maxsplit=2)
        env = splited_tag[1]
        if env == 'cmn': env = 'common'
        product = splited_tag[0]
        if product == 'dlp': product = 'delphi'
        if product == 'apl': product = 'apollo'
        if product == 'ph': product = 'prodhub'
        if product == 'wtlst': product = 'whitelist'

        next_frame = pd.DataFrame.from_dict({
            'year_month': pd.Series([date.rsplit('-', 1)[0] for date in row.keys()], dtype='str'),
            'month_cost': 0,
            'environment': env,
            'resources': splited_tag[2],
            'daily_cost': 0,
            'cost': pd.Series([round(x, MAX_DECIMAL_POINTS) for x in row.values()], dtype='float'),
            'date': pd.Series(row.keys(), dtype='str'),
            'product': product,
            'daily_product_cost': 0,
            'daily_environment_cost': 0,
        },)
        out_df = pd.concat([out_df, next_frame], ignore_index=True)
    return out_df

# Sort dataframe by date, product, environment
def sortBillingDf(in_df: pd.DataFrame)-> pd.DataFrame:
    out_df = in_df.sort_values(by=['date','product','environment'])
    return out_df

# Generate dataframe with summaries for monthly cost,
# daily cost, daily platform cost and daily platform environment cost
def generateSummaries(in_df: pd.DataFrame) -> pd.DataFrame:
    out_df = createEmptyDf()
    # Calculate sums for month, day, day_platform and day_platform_env
    for month in list(pd.unique(in_df['year_month'])):
        month_df = in_df.loc[in_df['year_month'] == month].copy(deep=False)
        month_cost = month_df['cost'].sum()
        month_df['month_cost']=month_cost
        for date in list(pd.unique(month_df['date'])):
            daily_df = month_df.loc[month_df['date'] == date].copy(deep=False)
            daily_cost = daily_df['cost'].sum()
            daily_df['daily_cost']=daily_cost

            for product in list(pd.unique(daily_df['product'].values)):
                daily_product_df = daily_df.loc[daily_df['product'] == product].copy(deep=False)
                daily_product_cost = daily_product_df['cost'].sum()
                daily_product_df['daily_product_cost'] = daily_product_cost

                for env in list(pd.unique(daily_product_df['environment'].values)):
                    daily_product_env_df = daily_product_df.loc[daily_df['environment'] == env].copy(deep=False)
                    daily_product_env_cost = daily_product_env_df['cost'].sum()
                    daily_product_env_df['daily_environment_cost'] = daily_product_env_cost
                    out_df = pd.concat([out_df, daily_product_env_df], ignore_index=True)
    return out_df