categorizer.py

import pandas as pd

CATEGORIES = {
    "DINING": ["RESTAURANT", "CAFE", "FOOD TRUCK", "STARBUCKS"],
    "TRANSPORT": ["UBER", "LYFT", "TAXI", "PETROL"],
    "SHOPPING": ["AMAZON", "WALMART", "TARGET"],
    "UTILITIES": ["ELECTRIC", "WATER", "INTERNET", "PG&E"]
}

def categorize_expenses(file):
    if isinstance(file, str):
        # Example file path
        df = pd.read_csv(file)
    else:
        # Uploaded file
        df = pd.read_csv(file.name)
    
    df['Category'] = "OTHER"
    
    for idx, row in df.iterrows():
        description = str(row['Description']).upper()
        for cat, keywords in CATEGORIES.items():
            if any(kw in description for kw in keywords):
                df.at[idx, 'Category'] = cat
                break
    return df