import gradio as gr
import pandas as pd
from transformers import pipeline

# Load a summarization model
summarizer = pipeline("summarization", model="facebook/bart-large-cnn")

def preprocess_csv(file_path):
    df = pd.read_csv(file_path)
    events = []
    for _, row in df.iterrows():
        events.append(f"On {row['Date']} at {row['Time']}, state was {row['State']} → {row['Message Text']}")
    return "\n".join(events)

def analyze_log(file_path, mode):
    text = preprocess_csv(file_path)

    # Build prompt depending on mode
    if mode == "Summarize":
        prompt = "Summarize the following AHU alarm log:\n\n" + text
    elif mode == "Highlight anomalies":
        prompt = "Identify unusual or repeated alarms in this AHU log and explain possible causes:\n\n" + text
    elif mode == "Suggest maintenance":
        prompt = "Based on this AHU alarm log, suggest maintenance actions:\n\n" + text
    else:
        prompt = text

    try:
        summary = summarizer(prompt, max_length=200, min_length=50, do_sample=False)
        return summary[0]['summary_text']
    except Exception as e:
        return f"Error during summarization: {e}"

iface = gr.Interface(
    fn=analyze_log,
    inputs=[
        gr.File(type="filepath", label="Upload Log File"),
        gr.Dropdown(choices=["Summarize", "Highlight anomalies", "Suggest maintenance"], label="Analysis Mode")
    ],
    outputs="text",
    title="AHU Log Analyzer",
    description="Upload your log file (CSV) and choose how you want it analyzed."
)

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