Python script violates the command-line character limit

I am using a Python script to batch convert many images in different folders into single pdfs (with https://pypi.org/project/img2pdf/):

import os
import subprocess
import img2pdf 
from shutil import copyfile

def main():
    folders = [name for name in os.listdir(".") if os.path.isdir(name)] 
    for f in folders:
        files = [f for f in os.listdir(f)] 
        p = ""
        for ffile in files:
            p += f+'\\'  + ffile + " "                
        os.system("py -m img2pdf *.pn* " + p + " --output " + f + "\combined.pdf")
    
        
if __name__ == '__main__':
    main()

However, despite running the command via Powershell on Windows 10, and despite using very short filenames, when the number of images is very high (eg over 600 or so), Powershell gives me the error "The command line is too long" and it does not create the pdf. I know there is a command-line string limitation (https://learn.microsoft.com/en-us/troubleshoot/windows-client/shell-experience/command-line-string-limitation), but I also know that for powershell this limit is higher (Powershell to avoid cmd 8191 character limit), and I can't figure out how to fix the script. I would like to ask you if you could help me fix the script to avoid violating the character limit. Thank you

PS: I use the script after inserting it in the parent folder that contains the folders with the images; then in each subfolder the output pdf file is created.

Using img2pdf library you can use this script:

import img2pdf
import os

for r, _, f in os.walk("."):
    imgs = []
    for fname in f:
        if fname.endswith(".jpg") or fname.endswith(".png"):
            imgs.append(os.path.join(r, fname))
    if len(imgs) > 0:
        with open(r+"\output.pdf","wb") as f:
            f.write(img2pdf.convert(imgs))