import requests
import sounddevice as sd
import base64
from transformers import pipeline

Initialize the translation pipeline

translator_en_to_fr = pipeline("translation_en_to_fr")
translator_fr_to_en = pipeline("translation_fr_to_en")

def translate_text(text, language):
"""
Translate text from one language to another.

Args:
    text (str): The text to translate.
    language (str): The language pair to use (e.g. "en_to_fr" or "fr_to_en").

Returns:
    str: The translated text.
"""
if language == "en_to_fr":
    return translator_en_to_fr(text)[0]['translation_text']
elif language == "fr_to_en":
    return translator_fr_to_en(text)[0]['translation_text']
else:
    raise ValueError("Invalid language pair")

def generate_audio(text, speaker_id, api_url, api_token):
"""
Generate audio for a given text using the Hugging Face API.

Args:
    text (str): The text to generate audio for.
    speaker_id (str): The speaker ID to use.
    api_url (str): The URL of the Hugging Face API.
    api_token (str): The API token to use.

Returns:
    tuple: A tuple containing the audio data and sampling rate.
"""
try:
    if not text or not speaker_id:
        raise ValueError("Text and speaker ID cannot be empty")
    headers = {
        "Authorization": f"Bearer {api_token}",
    }
    payload = {
        "inputs": text,
        "parameters": {
            "speaker_id": speaker_id
        }
    }
    response = requests.post(api_url, headers=headers, json=payload)
    response.raise_for_status()
    audio = response.json().get("audio")
    sampling_rate = response.json().get("sampling_rate")
    return audio, sampling_rate
except requests.exceptions.RequestException as e:
    print(f"Error: {e}")
    return None
except ValueError as e:
    print(f"Error: {e}")
    return None

def play_audio(audio, sampling_rate):
"""
Play audio using the sounddevice library.

Args:
    audio (bytes): The audio data to play.
    sampling_rate (int): The sampling rate of the audio.
"""
try:
    if not audio or not sampling_rate:
        raise ValueError("Audio and sampling rate cannot be empty")
    sd.play(audio, sampling_rate)
    sd.wait()
except sd.PortAudioError as e:
    print(f"Error: {e}")
except Exception as e:
    print(f"Error: {e}")

def main():
english_text = "Hello, my name is Eric Carnell Bearden and I like to talk."
french_text = "Bonjour, je m'appelle Eric Carnell Bearden et j'aime parler."
api_token = "YOUR_API_TOKEN" # Replace with your actual API token
speaker_id = "[spkr_63]"
api_url = "https://api-inference.huggingface.co/models/11mlabs/indri-0.1-350m-tts"

# English to French translation section
print("English to French translation:")
translated_french_text = translate_text(english_text, "en_to_fr")
print(f"English to French: {english_text} -> {translated_french_text}")

# Generate audio for English and French text
print("\nGenerating audio for English and French text:")
audio_en, sampling_rate_en = generate_audio(english_text, speaker_id, api_url, api_token)
if audio_en and sampling_rate_en:
    play_audio(audio_en, sampling_rate_en)

audio_fr, sampling_rate_fr = generate_audio(translated_french_text, speaker_id, api_url, api_token)
if audio_fr and sampling_rate_fr:
    play_audio(audio_fr, sampling_rate_fr)

# French to English translation section
print("\nFrench to English translation:")
translated_english_text = translate_text(french_text, "fr_to_en")