import pandas as pd
import matplotlib.pyplot as plt
import os

def read_smog_chamber_txt_files_as_dict(filename : str ,instrument_folder : str):

    
    if instrument_folder == 'smps':
        end_of_header = 'Sample #	Date	Start Time	Sample Temp (C)	Sample Pressure (kPa)'
        
    elif instrument_folder == 'gas':
        end_of_header = 'Date_Time	HoribaNO	HoribaNOy	Thermo42C_NO	Thermo42C_NOx	APHA370 CH4'
    else:
        raise ValueError('intrument_folder must be set as a either "smps" or "gas"')

    # Read header as a dictionary and detect where data table starts
    header_dict = {}
    data_start = False    
    with open(filename,'r') as f:
        for line_number, line in enumerate(f):        
            list = line.split('\t')
            if end_of_header in line:
                data_start = True  
                column_names = []
                for i, name in enumerate(line.split('\t')):
                    column_names.append(str(i)+'_'+name) 

                print(line_number, len(column_names ))
                break

            # TODO: update to extract information from lines formed by more than two elements separaed by '\t'
            key, value = list[0], list[1::]
            header_dict[key] = value

    if not data_start:
        raise ValueError('file appears to be invalid. Data start condition in txt file was not met.')
    
    df = pd.read_csv(filename, 
                     delimiter = "\t", 
                     header=line_number, 
                     #encoding='latin-1',
                     encoding='latin-1',
                     names=column_names,
                     skip_blank_lines=True)
    
    df_numerical_attrs = df.select_dtypes(include ='number')
    df_categorical_attrs = df.select_dtypes(exclude='number')

    if instrument_folder == 'smps':
        df_categorical_attrs['1_Timestamp'] = [ df_categorical_attrs.loc[i,'1_Date']+' '+df_categorical_attrs.loc[i,'2_Start Time'] for i in df.index]
        df_categorical_attrs = df_categorical_attrs.drop(columns=['1_Date','2_Start Time'])
    elif instrument_folder == 'gas':
        df_categorical_attrs = df_categorical_attrs.rename(columns={'0_Date_Time' : '0_Timestamp'})

    data_column_names = [item.encode("utf-8") for item in df_numerical_attrs.columns]
    output_dict = { 'header_dict':header_dict,
                    'data': df_numerical_attrs.to_numpy(),
                    'data_column_names':data_column_names,
                    'categ_data_dict':df_categorical_attrs.to_dict(orient='list')
                    }
    
   #output_dict = {'header_dict':header_dict,
   #                'num_data_df':df_numerical_attrs.to_numpy(),
   #                'categ_data_df':df_categorical_attrs.to_dict(orient='list')}

    return output_dict 

def main():



    #filename = 'M:\\gas\\20220705_000004_MSC_gases.txt' corrupted file
    #filename = 'M:\\gas\\20220726_101617_MSC_gases.txt'
    root_dir = '\\\\fs03\\Iron_Sulphate'

    instrument_folder = 'smps'
    filename_path = os.path.join(root_dir,'smps\\20220726\\20220726_num.TXT')


    instrument_folder = 'gas'
    filename_path = os.path.join(root_dir,'gas\\20220726_101617_MSC_gases.txt')
    


    result = read_smog_chamber_txt_files_as_dict(filename_path,instrument_folder)

    print(':)')

    return result


if __name__ == '__main__':

    output_dict  =  main()

    print(output_dict['num_data_df'].columns, len(output_dict['num_data_df'].columns),'\n')
    print(output_dict['num_data_df'].columns, len(output_dict['categ_data_df'].columns))