import math

class Element:
    id = ""
    count = 0
    p_x_i = 0
    I_x_i = 0

    def __init__(self, id):
        self.id = id
        self.count = 1

def find_element(list, target_it):
    for element in list:
        if element.id == target_it:
            return element
    return None

def read_text_file():
    path = './text.txt'

    with open(path, 'r', encoding='utf-8') as file:
        content = file.read()
    return content

def Z_statistik():
    # Ergebnisse mit https://de.planetcalc.com/2476/ prüfbar
    message = read_text_file()

    m = len(message)

    elements = []

    for char in message:
        entry = find_element(elements, char)
        if entry is not None:
            entry.count += 1
        else:
            elements.append(Element(char))

    calculate(elements, m, message)

def Z2_statistik():
    message = read_text_file()

    m = len(message) - 1        # Letztes Zeichen kann kein Tupel mit "nichts" bilden

    elements = []

    for i, char in enumerate(message):
        if i < m:
            tuple = char + message[i+1]
            entry = find_element(elements, tuple)

            if entry is not None:
                entry.count += 1
            else:
                elements.append(Element(tuple))

    calculate(elements, m, message)

def W_statistik():
    message = read_text_file()

    m = 0
    i = 0

    elements = []
    l = len(message)

    while i < l:
        word = ""

        if message[i].isalpha():
            # Fange ein Wort an
            word += message[i]

            # füge Zeichen hinzu bis Wortende erreicht ist
            while i+1 is not l and message[i+1].isalpha():
                i += 1
                word += message[i]

            # Erhöhe die Anzahl der Wörter in der Nachricht
            m += 1

            # Zähle die Häufigkeit der Wörter

            entry = find_element(elements, word)
            if entry is not None:
                entry.count += 1
            else:
                elements.append(Element(word))

        # gehe weiter in der Nachricht
        i += 1

    calculate(elements, m, message)

def calculate(elements, m, message):
    h = 0
    for element in elements:
        element.p_x_i = element.count / m
        element.I_x_i = -1 * math.log2(element.p_x_i)
        h += element.p_x_i * element.I_x_i

    # Ausgabe
    elements.sort(key = lambda element: element.count, reverse = True)
    for element in elements:
        print(f"{element.count:3.0f} | {element.p_x_i:10.7f} | {element.I_x_i:10.7f} | »{element.id}«")
    print("Entropie = " + h.__str__() + "\n\n")

def main():
    Z_statistik()
    Z2_statistik()
    W_statistik()

if __name__ == '__main__':
    main()