archive.extract "clean: True" cannot delete top level files

[angeloudy]

Description of Issue/Question

When apply archive.extract state, if clean is set to True, only files paths that belong to the zip file will be deleted.

For example, A zip file contains the following

test1.txt
test2.txt
testdir
testdir/test.txt

When applying archive.extract, everything in the list above will be deleted before the archive is extracted. However, if I add another file to the target directory test3.txt, it will never be deleted.

It's the same when you update your archive, say test2.txt is no longer part of the archive, test2.txt will never get deleted in the target directory.

This has caused some drama for us because we use archive.extract to deploy a lot of applications.

[angeloudy]

This could be resolved by always having a top level directory. But the problem with that is you may have to rename the director after extraction, which make the checksum useless.

[Ch3LL]

glancing at the code it looks like this is done on purpose. we could possibly add a new argument that deletes the files in the directory it extracts the archive too

[angeloudy]

@Ch3LL
That would be great.
If toplevel_dir is set, then ignore enforce_toplevel.

If there's more than one directory or any file exists in the top level, create toplevel_dir and then append toplevel_dir to name;
If there only one dir in the top level, rename the toplevel dir to toplevel_dir after extraction;

[Oloremo]

I think that is what you want: #55418

[Ch3LL]

ping @angeloudy can you verify the fix in #55418 looks like its ready to be closed.

[angeloudy]

@Ch3LL verified. Thanks for adding the new feature.