Python text to audio (Audio flash cards)

I want to begin learning a new language: Lets say French and there are many excellent resources available on platforms like YouTube. But these apps typically require an internet connection or other resources on a smartphone or laptop.

However, to make my language learning experience more effective and accessible, I am particularly focused on developing my listening skills. My goal is to keep things simple and I would like to use audio files that can be played on my compact MP3 player while I am commuting on job, walking or in gym.

Additionally, I would value the ability to update these audio files as I progress — whether it’s to add new content or remove outdated material. This flexibility would enable me to continually refine and personalize my learning experience, making the most of every opportunity to engage with the language learning process.

Selection of Tech Stack

For the sake of simplicity I have selected Python and found a fantastic library available and the library is gtts and what gtts is: gTTS (Google Text-to-Speech), a Python library and CLI tool to interface with Google text-to-speech translator API.

Get Sanjeev Tyagi’s stories in your inbox

Join Medium for free to get updates from this writer.

Remember me for faster sign in

To begin with, let’s create a file that will serve as the foundation for generating an audio file. For this purpose, I am using the JSON format and save the file with the name “misc.json.” as sample below (french : english) :

{
    "Je Suis":    "I am ",
    "Je suis la":    "I am here",
    "Je suis Pret":    "I m ready ",
    "Je suis content ":    "I am happy",
    "Je suis fatigue ":    "I am tired",
    "Je suis malade ":    "I am sick",
    "Je suis occupe ":    "I am busy"
}

Now, Just opened my terminal and installed gtts :

pip3 install gTTS

now writing script in file text_to_speech.py:

from gtts import gTTS
from gtts import lang

# Json format to provide text which need to convert in audio
import json

# command line argument 
import argparse

import sys
import os.path


if len(sys.argv) < 2:
    usage = """usage: text_to_speech: [-h] [-j JSON] [-l LANGUAGE] [-o [OPTIONS]] [-r REPETITION]

options:
  -h, --help            show this help message and exit
  -j JSON, --json JSON  Input Json file.
  -l LANGUAGE, --language LANGUAGE
                        second language
  -o OPTIONS, --options OPTIONS
                        List of language options available.
  -r REPETITION, --repetition REPETITION
                        Repeat word or statement of second language. Default is 3."""

    print(usage)
    exit()

args_parser=argparse.ArgumentParser(prog="text_to_speech:", description="""This script was originally created to generate MP3 audio for English and French text. It's mainly useful for listening practice.""")
args_parser.add_argument("-j", "--json", help="Input Json file.")

# default is French Language 
args_parser.add_argument("-l", "--language", default='fr', help="second language")
args_parser.add_argument("-o","--options", nargs='?', const=True, default=False, help="List of language options available.")

# default repetition is 3
args_parser.add_argument("-r", "--repetition", type=int, default=3, help="Repeat word or statement of second language. Default is 3.")


# parsing command line argument 
# like: Json file name 
# language (by default it is French)   
# Available languages listing opetion 
# and repetiton. default is 3
args=args_parser.parse_args()
if args.options:
    for short, language in lang.tts_langs().items():
        print(f"Laguage: {language} => short {short}")
    exit()

## create audio directory if not exists directory 
if not os.path.exists("audio"):
    os.mkdir("audio")

# extracting file name from json to generate audio file with same name 
fname = args.json.rsplit( ".", 1 )[0]
if not os.path.isfile("source"+ os.path.sep + args.json):
    print(f"Text source file '{args.json}' not exists in folder 'source' folder. Keep text json file in folder 'source' on same location as script")
    exit()

# Open source json file (sample as already provided  
with open("source" + os.path.sep +fname+".json", 'r') as f:
    french_words = json.load(f)

# generating mp3 audio file 
with open( "audio" + os.path.sep +fname+'.mp3', 'wb') as f:
    for k, v in french_words.items():

        # first option is english 
        lang = "en"
        gTTS( v + " .", lang=lang, ).write_to_fp(f)

        # second language is profided by command line            
        lang=args.language
        # with repetition from command line 
        for i in range(0, args.repetition):
            gTTS( k + " .", lang=lang, ).write_to_fp(f)
print( f"Audio file saved as audio{os.path.sep}{fname}.mp3")

# for the sake of simplicity, I am not adding exceptions.  

Now, final point. Execution of script

To get help:

python text_to_speech.py -h

## output would be 
usage: text_to_speech: [-h] [-j JSON] [-l LANGUAGE] [-o [OPTIONS]] [-r REPETITION]

options:
  -h, --help            show this help message and exit
  -j JSON, --json JSON  Input Json file.
  -l LANGUAGE, --language LANGUAGE
                        second language
  -o OPTIONS, --options OPTIONS
                        List of language options available.
  -r REPETITION, --repetition REPETITION
                        Repeat word or statement of second language. Default is 3."""

This script is designed to expect a JSON file located in the “source” folder, which should be under same directory as the script itself. Additionally, the output will be generated in a subfolder named “audio” within the same directory.

now, generate audio with sample file misc.json

python text_to_speech.py -j misc.json -l fr

this command will generate audio file as misc.mp3

Enjoy …