app.py

import os
import base64
import tempfile
from fastapi import FastAPI
from pydantic import BaseModel
from faster_whisper import WhisperModel
import gradio as gr
import uvicorn # <-- IMPORT THE SERVER

# --- 1. Configuration & Model Loading ---
os.environ["HF_HOME"] = "/tmp/huggingface_cache"
os.environ["HF_HUB_CACHE"] = "/tmp/huggingface_cache"
model = WhisperModel("Systran/faster-whisper-small", device="cpu", compute_type="int8")

# --- 2. FastAPI Application Setup ---
app = FastAPI()

class AudioInput(BaseModel):
    data: list[str]

def transcribe_audio(audio_filepath, language):
    if audio_filepath is None:
        return "Error: No audio file provided."
    lang = None if language == "auto" else language
    segments, _ = model.transcribe(audio_filepath, beam_size=5, language=lang, vad_filter=True)
    return " ".join(seg.text for seg in segments)

# --- 3. Create the API Endpoint ---
@app.post("/predict")
async def predict(audio_input: AudioInput):
    # The Gradio API sends data in a list, so we get the first item
    base64_data_uri = audio_input.data[0]
    
    # Handle the null test case from curl
    if base64_data_uri is None:
        return {"data": ["Error: No audio file provided."]}

    header, encoded_data = base64_data_uri.split(",", 1)
    audio_data = base64.b64decode(encoded_data)
    
    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as temp_audio_file:
        temp_audio_file.write(audio_data)
        temp_filepath = temp_audio_file.name
        
    try:
        transcription = transcribe_audio(temp_filepath, "auto")
    finally:
        os.remove(temp_filepath)
        
    return {"data": [transcription]}

# --- 4. Create the Gradio User Interface ---
iface = gr.Interface(
    fn=transcribe_audio,
    inputs=[
        gr.Audio(type="filepath", label="Upload Audio File"),
        gr.Radio(['en', 'bn', 'auto'], label="Select Language", value='auto')
    ],
    outputs="text",
    title="⚡ Zen Speech-to-Text (API Fixed)",
    description="Upload audio → get transcription"
)

# --- 5. Mount the Gradio UI onto the FastAPI App ---
app = gr.mount_gradio_app(app, iface, path="/")

# --- 6. Run the Server (THIS WAS THE MISSING PART) ---
if __name__ == "__main__":
    uvicorn.run(app, host="0.0.0.0", port=int(os.getenv("PORT", 7860)))