Merge PDFs

pdf_merge.py

from PyPDF2 import PdfMerger
import os


def merge_pdfs(input_folder, output_filename):
    """
    Recursively merge all PDFs in the specified folder and its subfolders into a single PDF file.

    Args:
        input_folder (str): Path to the folder containing PDF files
        output_filename (str): Name of the output merged PDF file
    """
    # Create a PdfMerger object
    merger = PdfMerger()

    # List to store all found PDF files with their full paths
    pdf_files = []

    # Recursively walk through directory
    for root, dirs, files in os.walk(input_folder):
        for file in files:
            if file.lower().endswith('.pdf'):
                # Get the full path of the PDF file
                file_path = os.path.join(root, file)
                pdf_files.append(file_path)

    # Sort files to ensure consistent ordering
    pdf_files.sort()

    # Check if there are any PDF files
    if not pdf_files:
        print("No PDF files found in the specified folder and its subfolders.")
        return

    try:
        # Append each PDF file to the merger
        for pdf_path in pdf_files:
            merger.append(pdf_path)
            # Get relative path for cleaner output
            relative_path = os.path.relpath(pdf_path, input_folder)
            print(f"Added: {relative_path}")

        # Write the merged PDF to the output file
        merger.write(output_filename)
        merger.close()

        print(f"\nSuccessfully merged {len(pdf_files)} PDF files into {output_filename}")

    except Exception as e:
        print(f"An error occurred: {str(e)}")

    finally:
        merger.close()


# Example usage
if __name__ == "__main__":
    # Replace these with your folder path and desired output filename
    folder_path = "addition_pages/addition-pages_56-70"
    output_file = f'{folder_path}.pdf'

    merge_pdfs(folder_path, output_file)