app.py

import gradio as gr
from langdetect import detect
from gtts import gTTS
import tempfile

# Define a mapping from language codes to language names.
lang_map = {
    'en': 'English',
    'es': 'Spanish',
    'fr': 'French',
    'de': 'German',
    'it': 'Italian',
    'pt': 'Portuguese',
    'nl': 'Dutch',
    'ru': 'Russian',
    'zh-cn': 'Chinese (Simplified)',
    'ja': 'Japanese',
    'ko': 'Korean',
    'pl': 'Polish',
    'uk': 'Ukrainian',
    'sk': 'Slovak',
    'lt': 'Lithuanian',
    'cs': 'Czech',
    'sr': 'Serbian',
    'hr': 'Croatian',
    'hi': 'Hindi'
}

def identify_and_pronounce(name, selected_lang):
    """
    This function detects the language of the given name and generates its pronunciation using gTTS.
    
    Parameters:
        name (str): The name input provided by the user.
        selected_lang (str): Either 'Auto' (to use detected language) or a specific language name to override.
        
    Returns:
        tuple: A status message string and the path to the generated audio file.
    """
    # Check for empty or whitespace-only name input.
    if not name or name.strip() == "":
        return "Please enter a name.", None

    # Attempt to detect the language of the input name.
    try:
        detected_lang = detect(name)
    except Exception as e:
        return f"Error detecting language: {str(e)}", None

    # Get the human-readable language name for the detected code.
    detected_lang_name = lang_map.get(detected_lang, 'English (default)')
    # Use English as default if detected language code is not in our mapping.
    detected_lang_code = detected_lang if detected_lang in lang_map else 'en'

    # Use the provided override language if it's not "Auto"
    if selected_lang != "Auto" and selected_lang in lang_map.values():
        lang_name = selected_lang
        # Get the corresponding language code for the selected language.
        lang_code = [code for code, name in lang_map.items() if name == selected_lang][0]
    else:
        lang_name = detected_lang_name
        lang_code = detected_lang_code

    # Generate pronunciation audio using gTTS.
    try:
        tts = gTTS(text=name, lang=lang_code, slow=False)
        # Create a temporary file in the allowed directory.
        # delete=False ensures the file persists after closing so that Gradio can serve it.
        temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
        tts.save(temp_file.name)  # Save the TTS output to the temporary file.
        temp_file.close()  # Close the file handle.
        return f"Detected language: {detected_lang_name}\nPronounced in: {lang_name}", temp_file.name
    except Exception as e:
        return f"Error generating pronunciation: {str(e)}", None

# Prepare language options for the dropdown input.
language_options = ["Auto"] + list(lang_map.values())

# Create a Gradio Interface.
# Note: Specify `type="filepath"` for gr.Audio so that it expects a file path.
interface = gr.Interface(
    fn=identify_and_pronounce,
    inputs=[
        gr.Textbox(label="Enter a name", value="Tomasz Durzyński"),
        gr.Dropdown(choices=language_options, label="Select Language (Auto uses detection)", value="Auto")
    ],
    outputs=[
        gr.Textbox(label="Language Info"),
        gr.Audio(label="Pronunciation", type="filepath")
    ],
    title="Name Language Detector and Pronouncer",
    description=("Enter a name to detect its language and hear it pronounced. "
                 "Optionally, select a language to override the default.")
)

# Launch the app.
# Bind to 0.0.0.0 so that HF Spaces can properly route requests to your app.
interface.launch(server_name="0.0.0.0")