import os   #compability with windows
from IPython import embed
import numpy as np

def parse_dataset(dataset_name):
    assert(os.path.exists(dataset_name))        #see if data exists
    f = open(dataset_name, 'r')                 #open data we gave in
    lines = f.readlines()                       #read data
    f.close()                                   #?

    # metadata lists for every loop
    eodfs = []
    deltafs = []
    stimulusfs = []

    # data itself
    times = []
    frequencies = []
    amplitudes = []

    # temporary lists with data we put in the lists above
    time = []
    ampl = []
    freq = []

    for i in range(len(lines)):
        l = lines[i].strip()                                        #all lines of textdata, exclude all empty lines (empty () default for spacebar)
        if "#" in l and "EODf" in l:                                #if line starts with # EODf:
            eodfs.append(float(l.split(':')[-1].strip()[:-2]))      #append: line splitted by ':' the 2nd part ([-1],
        if "#" in l and "Delta f" in l:                             #which got striped so we sure there is no space at the end,
            deltafs.append(float(l.split(':')[-1].strip()[:-2]))    #from that all expect the last two signs (Hz unit)
        if "#" in l and "StimulusFrequency" in l:                   #this for different metadata in different lists
            stimulusfs.append(float(l.split(':')[-1].strip()[:-2]))

        if '#Key' in l:
            #print('KEY')
            if len(time) != 0:              #therefore empty in the first round
                times.append(time)          #2nd loop means time != 0, so we put the times/amplitudes/frequencies to
                amplitudes.append(ampl)     #the data of the first loop
                frequencies.append(freq)

            time = []                       #temporary lists to overwrite the lists with the same name we made before
            ampl = []                       #so they are empty again
            freq = []
            print(len(times))

        if len(l) > 0 and l[0] is not '#':              #line not empty and doesnt start with #
            temporary = list(map(float, l.split()))     #temporary list where we got 3 index splitted by spacebar, map to find them
            time.append(temporary[0])                   #temporary lists with the data at that place, respectively
            freq.append(temporary[1])
            ampl.append(temporary[2])

    times.append(time)          #append data from one list to another
    amplitudes.append(ampl)     #these append the data from the first loop to the final lists, because we overwrite them (?)
    frequencies.append(freq)

    return times, frequencies, amplitudes, eodfs, deltafs, stimulusfs       #output of the function



def noise_reduce(dataset_name, n):
    assert (os.path.exists(dataset_name))  # see if data exists
    f = open(dataset_name, 'r')  # open data we gave in
    lines = f.readlines()  # read data
    f.close()
                                                #len of frequencies is 10 time shorter than before, so worked?
                                                #put in frequencies instead of dataset?
                                                #2nd loop cut frequencies by this function?
    cutf = []
    frequencies = []
    for i in range(len(lines)):
        l = lines[i].strip()

        if len(l) > 0 and l[0] is not '#':
            temporary = list(map(float, l.split()))
            frequencies.append(temporary[1])

    for k in np.arange(0, len(frequencies), n): # sollte nach k+n weitergehen?
        f = frequencies[k:k+n]
        mean = np.mean(f)
        cutf.append(mean)

    return cutf