pdf_merger_2.py

# pdf_merger_2.py
import os
from PyPDF2 import PdfMerger, PdfReader

def merge_pdfs_in_directory(input_folder, output_folder, output_filename="merged_output.pdf"):
    """
    Recursively finds all PDF files in the input_folder and merges them
    into a single PDF file in the output_folder.

    Args:
        input_folder (str): The path to the folder containing PDF files.
        output_folder (str): The path to the folder where the merged PDF will be saved.
        output_filename (str): The name of the merged PDF file.
    """
    if not os.path.exists(input_folder):
        print(f"Error: Input folder '{input_folder}' does not exist.")
        return

    # --- This is the part that creates the output directory if it doesn't exist ---
    if not os.path.exists(output_folder):
        os.makedirs(output_folder)
        print(f"Created output folder: '{output_folder}'")
    # -----------------------------------------------------------------------------

    merger = PdfMerger()
    pdf_files_found = []

    # Walk through the input folder recursively
    for root, _, files in os.walk(input_folder):
        for file in files:
            if file.lower().endswith('.pdf'):
                filepath = os.path.join(root, file)
                pdf_files_found.append(filepath)

    if not pdf_files_found:
        print(f"No PDF files found in '{input_folder}' and its subdirectories.")
        return

    # Sort the files for consistent merging order (optional, but good practice)
    pdf_files_found.sort()

    print(f"Found {len(pdf_files_found)} PDF files to merge:")
    for pdf_file in pdf_files_found:
        print(f"- {pdf_file}")
        try:
            with open(pdf_file, 'rb') as f:
                merger.append(PdfReader(f))
        except Exception as e:
            print(f"Error appending {pdf_file}: {e}")

    output_filepath = os.path.join(output_folder, output_filename)

    try:
        with open(output_filepath, 'wb') as output_file:
            merger.write(output_file)
        print(f"\nSuccessfully merged PDF files to: '{output_filepath}'")
    except Exception as e:
        print(f"Error writing merged file: {e}")
    finally:
        merger.close()

if __name__ == "__main__":
    script_dir = os.path.dirname(os.path.abspath(__file__))
    input_folder_name = "input"
    output_folder_name = "output"

    # Construct full paths
    input_path = os.path.join(script_dir, input_folder_name)
    output_path = os.path.join(script_dir, output_folder_name)

    # Example Usage:
    # 1. Create an 'input' folder in the same directory as this script.
    # 2. Place some PDF files (and even subfolders with PDFs) inside the 'input' folder.
    # 3. Run this script.
    # The merged PDF will be created in the 'output' folder.
    merge_pdfs_in_directory(input_path, output_path, "all_merged_documents.pdf")