Import grouped data in a CSV into Pandas and substitute NaNs through real data

import_grouped_data.py

import csv

reader = csv.reader(open('products.csv'))
products_list = []
first_of_group = ""

for row in reader:
    product_name, feature1, feature2 = row
    product_dict = {}
    
    if product_name == "":
        product_dict["product_name"] = first_of_group
    else:
        product_dict["product_name"] = product_name
        first_of_group = product_name
        
    product_dict["feature1"] = feature1
    product_dict["feature2"] = feature2
    products_list.append(product_dict)

products_df = pd.DataFrame(products_list)[["product_name", "feature1", "feature2"]]
products_df

products.csv