How to create mp3 from text with Python – updated version

create mp3 from text – new version of textspeak
# create the mp3 letters audio files
import os
from gtts import gTTS
import tkinter as tk
from pygame import mixer
import tkinter.ttk as ttk
from time import sleep

if "audio" not in os.listdir():
    os.mkdir("audio")

def speak(word):
    print("looking for word in dictionary")
    # print(os.listdir("audio/"))
    word = word.lower()
    if f"{word}.mp3" not in os.listdir("audio/"):
        if word != "":
            print("This is not present... creating")
            t = gTTS(word, "it")
            t.save(f"audio/{word}.mp3")
            sleep(1)
            play(word)
            print("created " + word)
            show_list()
            lbx.focus(word)
    else:
        play(word)


def show_list():
    global lbx

    lbx.delete("0", "end")
    elenco = os.listdir("audio")
    elenco.sort(reverse=False)
    for file in elenco:
        if len(file[:-4]) == 1:
            lbx.insert("end", f"")
            lbx.insert("end", f"=== [ {file[:-4]} ] ====")
        else:
            lbx.insert("end", file)

def delete(evt):
    "with control+d it deletes the sound selected"
    global Lbx

    name = lbx.get(lbx.curselection())
    print(name)
    os.remove("audio/" + name)
    lbx.delete("anchor")

def play(word):
    mixer.music.load("audio/" + word + ".mp3")
    mixer.music.play()

def start():
    global word, lbx, text

    mixer.init()
    word = ""
    def audiokey(event):
        global word

        word = word.lower()
        print(event.char)
        # mixer.music.load("audio/" + str(event.char) + ".mp3")
        # mixer.music.play()
        try:
            if event.char  in " ,.!?\n":
                speak(word)
                play(word)
                word = ""
            else:
                word += str(event.char)

        except:
            word=""

    def indietro(event):
        global word
        
        word = word[:-1]
      
    root = tk.Tk()
    root.title("Type to hear the pc reading the text when press space")
    root.geometry("600x400+200+400")
    root.config(bg="coral")
    # text widget where you write
    lbx = tk.Listbox(
        root,
        bg="black",
        fg="white"
        )
    lbx.pack(side="left", fill="both")
    lbx.bind("<Double-Button>", add_word_to_text)
    lbx.bind("<Control-d>", delete)
    scrollbar = ttk.Scrollbar(root, orient=tk.VERTICAL, command=lbx.yview)
    scrollbar.pack(side="left", fill=tk.Y)
    lbx.config(yscrollcommand=scrollbar.set)
    show_list()

    lab = tk.Label(root, text="Write a full sentence and press return", bg="coral")
    lab.pack(fill="x")

    e = tk.Entry(root, bg="black", fg="white",
        insertbackground="white")
    e.pack(fill="both", expand=True)
    e.bind("<Return>", lambda evt: save_text(e.get()))
    
    lab = tk.Label(root, text="When you press space an mp3 is generated in the folder audio",
      bg="coral")
    lab.pack()

    text = tk.Text(root, height=15, bg="black", fg="white",
        insertbackground="white")
    text.pack()

    button = tk.Button(root, text="Apri cartella con i file audio / open audio folder",
        command=openfile, bg="black", fg="white",)
    button.pack()

    text.focus()
    text.bind("<Key>", audiokey)
    text.bind("<BackSpace>", indietro)
    # text.bind("<Return>", _return)
    root.mainloop()


def openfile():
    os.system("start audio\\")

def save_text(text):
    speak(text)

def add_word_to_text(evt):
    global lbx, text

    word = lbx.get(lbx.curselection())
    word = word[:-4]
    text.insert("end", word)
    mixer.music.load("audio/" + word + ".mp3")
    mixer.music.play()
    text.insert("end", " ")


start()
https://github.com/formazione/textspeak