app.py

import streamlit as st
import pandas as pd
from PIL import Image, ImageDraw, ImageFont
import io

def main():
    # Sidebar for dropdown, buttons, and instructions
    st.sidebar.title("AI Energy Score Label Generator")
    st.sidebar.write("### Instructions:")
    st.sidebar.write("1. Select a model from the dropdown.")
    st.sidebar.write("2. Review the label preview.")
    st.sidebar.write("3. Download the label as a PNG.")
    st.sidebar.markdown("[Learn more about AI Energy Scores](https://example.com)")

    # Read Data from CSV
    try:
        data_df = pd.read_csv("data.csv")
    except FileNotFoundError:
        st.sidebar.error("Could not find 'data.csv'! Please make sure it's present.")
        return

    # Ensure the CSV has required columns
    required_columns = ["model", "provider", "date", "task", "hardware", "energy", "score"]
    for col in required_columns:
        if col not in data_df.columns:
            st.sidebar.error(f"The CSV file must contain a column named '{col}'.")
            return

    # Dropdown for selecting a model
    model_options = data_df["model"].unique().tolist()
    selected_model = st.sidebar.selectbox("Select a Model:", model_options)

    # Filter the data for the selected model
    model_data = data_df[data_df["model"] == selected_model].iloc[0]

    # Dynamically select the background image based on the score
try:
    score = int(model_data["score"])  # Convert to int
    background_path = f"{score}.png"  # E.g., "1.png", "2.png"
    background = Image.open(background_path).convert("RGBA")

    # Proportional scaling to fit within the target size
    target_size = (800, 600)  # Maximum width and height
    background.thumbnail(target_size, Image.ANTIALIAS)

except FileNotFoundError:
    st.sidebar.error(f"Could not find background image '{score}.png'. Using default background.")
    background = Image.open("default_background.png").convert("RGBA")
    background.thumbnail(target_size, Image.ANTIALIAS)  # Resize default image proportionally
except ValueError:
    st.sidebar.error(f"Invalid score '{model_data['score']}'. Score must be an integer.")
    return

    # Generate the label with text
    generated_label = create_label(background, model_data)

    # Display the label
    st.image(generated_label, caption="Generated Label Preview")

    # Download button for the label
    img_buffer = io.BytesIO()
    generated_label.save(img_buffer, format="PNG")
    img_buffer.seek(0)

    st.sidebar.download_button(
        label="Download Label as PNG",
        data=img_buffer,
        file_name="AIEnergyScore.png",
        mime="image/png"
    )

def create_label(background_image, model_data):
    """
    Create the label image by adding text from model_data to the background image.
    """
    label_img = background_image.convert("RGBA")
    draw = ImageDraw.Draw(label_img)

    # Load the Inter variable font (no LAYOUT_RAQM)
    try:
        inter_font_path = "Inter-VariableFont_opsz,wght.ttf"
        title_font = ImageFont.truetype(inter_font_path, 16)  # Bold for title
        details_font = ImageFont.truetype(inter_font_path, 12)  # Medium for details
        energy_font = ImageFont.truetype(inter_font_path, 14)  # Medium for energy
    except Exception as e:
        st.error(f"Font loading failed: {e}")
        return label_img

    # Define positions for each text group
    title_x, title_y = 100, 120
    details_x, details_y = 200, 340
    energy_x, energy_y = 600, 600

    # Group 1: Title (Left-Justified, no prefixes)
    draw.text((title_x, title_y), model_data['model'], font=title_font, fill="black")
    draw.text((title_x, title_y + 25), model_data['provider'], font=title_font, fill="black")

    # Group 2: Details (Right-Justified, no prefixes)
    details_lines = [
        model_data['date'],
        model_data['task'],
        model_data['hardware']
    ]
    for i, line in enumerate(details_lines):
        bbox = draw.textbbox((0, 0), line, font=details_font)
        text_width = bbox[2] - bbox[0]
        draw.text((details_x - text_width, details_y + i * 20), line, font=details_font, fill="black")

    # Group 3: Energy (Bottom-Center, no prefixes)
    energy_text = model_data['energy']
    bbox = draw.textbbox((0, 0), energy_text, font=energy_font)
    energy_text_width = bbox[2] - bbox[0]
    draw.text((energy_x - energy_text_width // 2, energy_y), energy_text, font=energy_font, fill="black")

    return label_img

if __name__ == "__main__":
    main()