import gradio as gr
from transformers import pipeline

# Load Whisper for speech-to-text
whisper = pipeline("automatic-speech-recognition", model="openai/whisper-medium")

# Load a sentiment analysis model
sentiment_analyzer = pipeline("sentiment-analysis", model="cardiffnlp/twitter-roberta-base-sentiment")

# Function to process audio and analyze tone
def analyze_call(audio_file):
    try:
        # Step 1: Transcribe audio to text using Whisper
        transcription = whisper(audio_file)["text"]
        
        # Step 2: Analyze sentiment of the transcription
        sentiment_result = sentiment_analyzer(transcription)[0]
        
        # Prepare the output
        output = {
            "transcription": transcription,
            "sentiment": sentiment_result["label"],
            "confidence": round(sentiment_result["score"], 4)
        }
        return output
    except Exception as e:
        return {"error": str(e)}

# Gradio Interface
def gradio_interface(audio):
    if audio is None:
        return "Please record or upload an audio file."
    result = analyze_call(audio)
    if "error" in result:
        return f"Error: {result['error']}"
    return (
        f"**Transcription:** {result['transcription']}\n\n"
        f"**Sentiment:** {result['sentiment']}\n\n"
        f"**Confidence:** {result['confidence']}"
    )

# Create Gradio app
interface = gr.Interface(
    fn=gradio_interface,
    inputs=gr.Audio(type="filepath", label="Record or Upload Audio"),
    outputs=gr.Textbox(label="Analysis Result", lines=5),
    title="Real-Time Call Analysis",
    description="Record or upload audio to analyze tone and sentiment in real time.",
    live=False  # Set to False to avoid constant re-runs
)

# Launch the app
interface.launch()