Python Practice 18/22

Fix the bug: the list shrinks while you walk it

Every deploy runs the folder through clean_files first. Editors leave .bak copies of whatever you were last editing, and a web server does not know what a .bak is - it just hands the raw file to anyone who asks for it by name. So the script strips them before upload, and it has been stripping them for a year without complaint.

Last Friday styles.css.bak showed up on the live site.

Nothing crashed. No traceback, no warning, no log line. clean_files was handed six filenames, it removed two of the three backups, and it returned a list that looks perfectly reasonable until you count what is in it. Run it and read the output against the input - the survivor is sitting right there.

The loop is fine. The list is fine. What breaks is doing both at once.

Your task: fix clean_files(files) so it returns every filename that does not end in .bak - all three backups gone, every other file kept, in order.

You'll practice:

Reading a returned list as a claim, and counting whether the claim holds
Spotting a loop whose list is being shortened underneath it

# clean_files should drop every .bak file before upload. One of them survives.
def clean_files(files):
    for f in files:
        if f.endswith(".bak"):
            files.remove(f)
    return files

print(clean_files(["app.js", "app.js.bak", "index.html", "logo.png", "notes.txt.bak", "styles.css.bak"]))