app.py

import gradio as gr
import requests
import numpy as np
import io
import soundfile as sf  # To read audio data into NumPy array

# Placeholder API URLs (Replace with actual endpoints)
TRANSLATION_API_URL = "https://twcc2.eztalking.ai/nantrans/inference"
TTS_API_URL = "http://twcc2.eztalking.ai/mtts/tts"


def fetch_translation(text, target_language):
    # Mock implementation (Replace with actual API call)
    payload = {
        "input_text": text,
        "id": "1",
        "src_lang": "zh",
        "tgt_lang": "tw"
    }

    response = requests.post(TRANSLATION_API_URL, json=payload)
    if response.status_code == 200:
        return response.text
    return "Translation failed."


def fetch_tts_audio(translated_text, spk):
    # Mock implementation (Replace with actual API call)
    payload = {
        "input_text": translated_text,
        "id": "1",
        "src_lang": "tw",
        "tgt_lang": "tailo"
    }

    # response = requests.post(TRANSLATION_API_URL, json=payload)
    # if response.status_code == 200:
    #     translated_text = response.text
    payload = {
        "text": translated_text,  # "tw_convert": False,
        "b64enc": False, "speaker": spk, "speed": 0.9
    }

    response = requests.post(TTS_API_URL, json=payload)
    if response.status_code == 200:
        # Read the audio data from the response content into a NumPy array
        audio_data, sample_rate = sf.read(io.BytesIO(response.content))
        return audio_data, sample_rate
    return None, None


def translate_and_speak(text, target_language, spk):
    translated_text = fetch_translation(text, target_language)

    if not translated_text:
        return "Translation failed.", None

    audio_data, sample_rate = fetch_tts_audio(translated_text, spk)
    if audio_data is not None:
        return translated_text, (sample_rate, audio_data)
    return translated_text, None


spk_list = requests.get("http://twcc2.eztalking.ai/mtts/list_speakers").json()

# Gradio Interface
iface = gr.Interface(
    fn=translate_and_speak,
    inputs=[
        gr.Textbox(label="Enter Text"),
        gr.Dropdown(choices=["en", "es", "fr", "de"], label="Target Language", value="en"),
        gr.Dropdown(choices=spk_list, label="Select Speaker", value=spk_list[0])
    ],
    outputs=[
        gr.Textbox(label="Translated Text"),
        gr.Audio(label="TTS Audio", type="numpy")
    ],
    title="Text Translator with TTS",
    description="Translate text to a selected language and generate TTS audio."
)

iface.launch()