GUI'less Highlight Extraction

import csv
from docx import Document

# Map highlight colors to actors
COLOR_TO_ACTOR = {
    "YELLOW": "NARRATOR",
    "GREEN": "ACTOR 1",
    "RED": "Cade"
}

def extract_highlighted_text_with_order(file_path):
    """
    Extract highlighted text, map it to actors, and add order and color information.
    :param file_path: Path to the Word document.
    :return: List of dictionaries with 'Order', 'Actor', 'Color', and 'Line'.
    """
    document = Document(file_path)
    extracted_data = []
    order = 1

    for paragraph in document.paragraphs:
        for run in paragraph.runs:
            if run.font.highlight_color:
                highlight_color = run.font.highlight_color.name  # Use .name to get the color name
                actor = COLOR_TO_ACTOR.get(highlight_color, "Unknown")
                extracted_data.append({
                    "Order": order,
                    "Actor": actor,
                    "Color": highlight_color,
                    "Line": run.text.strip()
                })
                order += 1

    return extracted_data

def save_to_csv(data, output_file):
    """
    Save extracted data to a CSV file.
    :param data: List of dictionaries with extracted data.
    :param output_file: Path to the output CSV file.
    """
    with open(output_file, mode="w", newline="", encoding="utf-8") as csv_file:
        writer = csv.DictWriter(csv_file, fieldnames=["Order", "Actor", "Color", "Line"])
        writer.writeheader()
        writer.writerows(data)

if __name__ == "__main__":
    # Path to your Word document
    input_file = "Python\Personal\Highlighted_document_extractor\Document.docx"  # Replace with your document path
    output_file = "Python\Personal\Highlighted_document_extractor\output.csv"  # Replace with desired output CSV path

    # Extract highlighted text
    highlighted_text = extract_highlighted_text_with_order(input_file)

    # Save to CSV
    save_to_csv(highlighted_text, output_file)

    print(f"Extracted data has been saved to {output_file}.")