ui_core.py

import os
import logging
import json
import pickle
import numpy as np
from typing import Tuple, List, Dict, Any, Optional
import gradio as gr

# Configure logging
logging.basicConfig(
    level=logging.INFO,
    format='%(asctime)s - %(name)s - %(levelname)s - %(message)s'
)
logger = logging.getLogger('ui_core')

# Global variables
embeddings = {}
# Update default path to point to the pickle file
EMBEDDINGS_PATH = os.environ.get('EMBEDDINGS_PATH', 'data/ingredient_embeddings_voyageai.pkl')
CATEGORY_EMBEDDINGS_PATH = os.environ.get('CATEGORY_EMBEDDINGS_PATH', 'data/category_embeddings.pickle')

def load_embeddings(filepath: str = EMBEDDINGS_PATH) -> Dict[str, Any]:
    """
    Load ingredient embeddings from a pickle file
    
    Args:
        filepath: Path to the embeddings file
        
    Returns:
        Dictionary of ingredient embeddings
    """
    try:
        logger.info(f"Attempting to load embeddings from: {filepath}")
        if not os.path.exists(filepath):
            logger.error(f"Embeddings file not found: {filepath}")
            # Try alternative file formats
            alt_paths = [
                filepath.replace('.pkl', '.pickle'),
                filepath.replace('.pickle', '.pkl'),
                'data/ingredient_embeddings_voyageai.pkl',
                'data/ingredient_embeddings.pickle'
            ]
            for alt_path in alt_paths:
                if os.path.exists(alt_path) and alt_path != filepath:
                    logger.info(f"Found alternative embeddings file: {alt_path}")
                    filepath = alt_path
                    break
            else:
                return {}
            
        # Determine file type and load accordingly
        if filepath.endswith(('.pkl', '.pickle')):
            logger.info(f"Loading pickle file: {filepath}")
            with open(filepath, 'rb') as f:
                loaded_embeddings = pickle.load(f)
        else:
            logger.info(f"Loading JSON file: {filepath}")
            with open(filepath, 'r') as f:
                loaded_embeddings = json.load(f)
            
        # Validate the loaded data
        if not isinstance(loaded_embeddings, dict) or not loaded_embeddings:
            logger.error(f"Invalid embeddings format in {filepath}")
            return {}
            
        # Convert lists to numpy arrays for faster processing
        processed_embeddings = {}
        for ingredient, embedding in loaded_embeddings.items():
            if isinstance(embedding, list):
                processed_embeddings[ingredient] = np.array(embedding)
            else:
                processed_embeddings[ingredient] = embedding
            
        logger.info(f"Successfully loaded {len(processed_embeddings)} ingredient embeddings")
        return processed_embeddings
        
    except json.JSONDecodeError:
        logger.error(f"Invalid JSON format in embeddings file: {filepath}")
        return {}
    except pickle.UnpicklingError:
        logger.error(f"Invalid pickle format in embeddings file: {filepath}")
        return {}
    except Exception as e:
        logger.error(f"Error loading embeddings: {str(e)}")
        return {}

# Load embeddings at module import time
embeddings = load_embeddings()

# If embeddings is empty, try loading category embeddings
if not embeddings:
    logger.info("No ingredient embeddings found, trying category embeddings...")
    embeddings = load_embeddings(CATEGORY_EMBEDDINGS_PATH)

# Sample product names for the example button
EXAMPLE_PRODUCTS = """Nature's Promise Spring Water Multipack
Red's Burritos
Nature's Promise Spring Water Multipack
Schweppes Seltzer 12 Pack
Hunt's Pasta Sauce
Buitoni Filled Pasta
Buitoni Filled Pasta
Samuel Adams or Blue Moon 12 Pack
Mrs. T's Pierogies
Buitoni Filled Pasta
Pillsbury Dough
Nature's Promise Organic Celery Hearts
MorningStar Farms Meatless Nuggets, Patties or Crumbles
Nature's Promise Organic Celery Hearts
Boar's Head Mild Provolone Cheese
Athenos Feta Crumbles"""

def load_examples():
    """Load example product names into the text input"""
    return EXAMPLE_PRODUCTS

# Removed obsolete theme/CSS imports and functions previously here.
# Theme is now handled by Streamlit via .streamlit/config.toml

def parse_input(input_text, is_file=False) -> Tuple[List[str], Optional[str]]:
    """Parse user input into a list of product names"""
    try:
        if is_file:
            # Handle file input (assuming newline-separated product names)
            product_names = [line.strip() for line in input_text.split('\n') if line.strip()]
        else:
            # Handle text input (assuming newline-separated product names)
            product_names = [line.strip() for line in input_text.split('\n') if line.strip()]
            
        if not product_names:
            return [], "<div style='color: #d32f2f; font-weight: bold; padding: 20px;'>No valid product names found. Please check your input.</div>"
            
        return product_names, None
        
    except Exception as e:
        logger.error(f"Error parsing input: {str(e)}")
        return [], f"<div style='color: #d32f2f; font-weight: bold; padding: 20px;'>Error parsing input: {str(e)}</div>"