trailing slashes for "leaf" directories, usage of '\\'

[h5rdly]

Hi,

I Wanted to ask regarding some parsing issues I'm having.

On Windows, Zipfile listings look like this:
'Folder/Folder2/FILE'
'Folder/Folder2/EmptyFolder/'

With Rarfile it's:
u'Folder\\Folder2\\FILE'
u'Folder\\Folder2\\EmptyFolder'

This causes obstacles -

1. With the lack of trailing slashes, I need extra info to know if FILE a file or a folder.
2. I would need to parse windows and linux rars differently, as well as rarfiles and zipfiles, gzipfiles etc.

Would it be possible to add trailing slashes to last-in-line folders, and/or use Python's generic platform agnostic '/' instead of '\\'? That would be awesome, as it would allow to utilize a single generic archive parsing.

Thanks.

[XenHat]

The option to use a different path separator already exists. Did you take a look at the documentation linked in the README file?
The option you are looking for is called PATH_SEP, see https://rarfile.readthedocs.io/en/latest/api.html#rarfile.PATH_SEP

Edit: about 1.; You can use os.path.isdir in Python. See https://docs.python.org/3/library/os.path.html#os.path.isdir for documentation, replace /3/ with /2/ in the URL for python 2.x

[markokr]

Actually in rarfile 3.0 the default path_sep is '/' now.

But rarfile does not have any special handling for dirs, you are supposed to check mode.

Seems ZIP format stores final '/' in archive, while RAR does not. Should rarfile append '/' to dirs?

[h5rdly]

@XenHat
I was not aware of rarfile.PATH_SEP = '/', thanks for the reference.

is_dir() isn't much use for traversing the archive, like in a file manager. One can get the dir info either way, but it's an extra check on every "leaf" folder, might add up on vast archives so it's a bit of a shame. Also not compatible with the Zipfile style.

@markokr
I'm obviously biased, but it would make parsing smooth, as it would just need the pathname to figure dirs from files.
Even looking at a listing, it's sometimes hard to tell if something like "THIS" or "notes" is a file or a directory.
The question is really, why not? :)

[XenHat]

Seems ZIP format stores final '/' in archive, while RAR does not. Should rarfile append '/' to dirs?

Since the RARFile information says it tries to behave as closely to zipfile as possible, I would say yes.

Footnote: I wonder if the lack of trailing slash would cause my odd problem with some archives (#29 )since it tends to happen with archives with multiple subfolder levels (the most common example I have is libboost, which is huge, tree-wise)?