app.py

import gradio as gr
import os
import subprocess
from urllib.request import urlretrieve
import traceback

# --- Create a directory for the model files ---
os.makedirs("tts_model", exist_ok=True)

# --- Model and Config URLs ---
MODEL_URL = "https://huggingface.co/SYSPIN/vits_Chhattisgarhi_Female/resolve/main/best_model.pth?download=true"
CONFIG_URL = "https://huggingface.co/SYSPIN/vits_Chhattisgarhi_Female/resolve/main/config.json?download=true"

# --- Define local paths for the model and config ---
model_path = os.path.join("tts_model", "best_model.pth")
config_path = os.path.join("tts_model", "config.json")

# --- Download Model and Config if they don't exist ---
def download_file(url, destination):
    if not os.path.exists(destination):
        print(f"Downloading {os.path.basename(destination)}...")
        try:
            urlretrieve(url, destination)
            print("Download complete.")
        except Exception as e:
            print(f"FATAL: Error downloading {url}: {e}")
            raise

try:
    download_file(MODEL_URL, model_path)
    download_file(CONFIG_URL, config_path)
except Exception as e:
    # If downloads fail, we cannot proceed.
    print("Application cannot start because model files could not be downloaded.")
    # This will cause the app to crash, which is appropriate here.
    raise

def chhattisgarhi_tts_cli(text):
    """
    Converts Chhattisgarhi text to speech using the TTS command-line interface.
    Returns a tuple: (status_message, audio_filepath)
    """
    if not text.strip():
        return ("Warning: Please enter some text to convert.", None)

    try:
        output_wav_path = "output.wav"
        
        # Construct the command-line instruction
        command = [
            "tts",
            "--text", text,
            "--model_path", model_path,
            "--config_path", config_path,
            "--out_path", output_wav_path
        ]
        
        print(f"Running command: {' '.join(command)}")

        # Execute the command. Capture stderr for detailed error reporting.
        result = subprocess.run(
            command, 
            check=True,        # Raise an exception if the command fails
            capture_output=True, 
            text=True
        )

        print(f"TTS command successful. Output saved to {output_wav_path}")
        return ("Success!", output_wav_path)

    except subprocess.CalledProcessError as e:
        # This block is executed if the 'tts' command returns a non-zero exit code (i.e., an error)
        error_message = f"TTS command failed."
        # The actual error from the TTS tool is in stderr
        stderr_output = e.stderr.strip()
        print(f"ERROR: {error_message}")
        print(f"STDERR from TTS tool: {stderr_output}")
        # Return a user-friendly error message to the Gradio interface
        return (f"Error during audio generation. Details: {stderr_output}", None)
        
    except Exception as e:
        # Catch any other unexpected errors (e.g., file not found if downloads failed silently)
        error_message = f"An unexpected error occurred: {str(e)}"
        print(error_message)
        traceback.print_exc()
        return (error_message, None)

# --- Gradio Interface ---
title = "Chhattisgarhi TTS Conversion"
description = """
### Chhattisgarhi Text-to-Speech
Enter text in Chhattisgarhi to generate audio. This app uses the `tts` command-line tool to run a VITS model 
provided by [SYSPIN on Hugging Face](https://huggingface.co/SYSPIN/vits_Chhattisgarhi_Female).
"""

examples = [
    ["राजस्थान के नामी ब्यंजन चूरमालाड़ू गुड़ के पाग म गहूँ के दरदरहा पिसान के लाड़ू म तिली अउ नरियल के सुवाद म सजथे"],
    ["दुग्ध क्रान्ति भारत के योजना हे जेखर ले भारत म दूध के कमी ला दुरिहा करे जा सकथे एला श्वेत क्रांति घलोक कहिथे"],
    ["जम्मू कश्मीर म पर्यटन उद्योग ला बढ़ावा देना उहाँ के अर्थबेवस्था ला सुचारू रूप ले चलाय बर जरुरी हे"],
    ["फोरेन्सिक विज्ञान ल कानूनी प्रश्न के उत्तर दिए बर अउ अपराध ल सिद्ध करे बर बउरे जाथे"]
]

iface = gr.Interface(
    fn=chhattisgarhi_tts_cli,
    inputs=gr.Textbox(lines=3, label="Enter Chhattisgarhi Text", placeholder="यहाँ छत्तीसगढ़ी पाठ लिखें..."),
    outputs=[
        gr.Textbox(label="Status", interactive=False),
        gr.Audio(type="filepath", label="Generated Speech")
    ],
    title=title,
    description=description,
    examples=examples,
    allow_flagging="never",
    cache_examples=True 
)

if __name__ == "__main__":
    iface.launch()