import re

header = """
@RELATION adtracking


@ATTRIBUTE ip NUMERIC
@ATTRIBUTE app NUMERIC
@ATTRIBUTE device NUMERIC
@ATTRIBUTE os NUMERIC
@ATTRIBUTE channel NUMERIC
@ATTRIBUTE click_time_y NUMERIC
@ATTRIBUTE click_time_m NUMERIC
@ATTRIBUTE click_time_d NUMERIC
@ATTRIBUTE is_attributed {0, 1}


@DATA

"""


def separated_date(lista):
    date = lista[:3]
    horario = lista[3:]
    date_string = ""
    for el in date:
        date_string += el + ","

    horario_string = ""
    for el in horario:
        horario_string += el + ","
    return date_string + horario_string



def resultset_toString(lista):
    date = lista[:3]
    horario = lista[3:]
    date_string = ""
    for el in date:
        date_string += el + "-"

    horario_string = ""
    for el in horario:
        horario_string += el + ":"

    return """'""" + date_string[:-1] + " " + horario_string[:-1] + """'"""

def format_line(linez):
    one_date = False
    result = re.findall(r'(\d{4})-(\d{2})-(\d{2}) (\d{2}):(\d{2}):(\d{2})', linez)
    if len(result) <= 0:
        return linez
    dates_formated = []
    for i in result:
        dates_formated.append(resultset_toString(i))

    if(len(dates_formated) <= 1):
        one_date = True
    dates_decomposited = []

    for i in result:
        dates_decomposited.append(separated_date(i))

    for el in list(zip(dates_formated, dates_decomposited)):
        data_formatada = el[0][:-1]
        data_formatada = data_formatada[1:]
        if(one_date):
            linez = linez.replace(data_formatada, el[1])
            linez = linez.replace(",,", "")
        else:
            linez = linez.replace(data_formatada, el[1])
    return linez

fname = "train_sample.csv"

file_to_write = open("test1.csv", 'a')
file_to_write.write(header)
with open(fname) as file_to_read:
    for line in file_to_read:
        file_to_write.write(format_line(line))