[Python-Dev] Updates to PEP 471, the os.scandir() proposal

[Victor Stinner]

2014-07-08 22:09 GMT+02:00 Ben Hoyt <benhoyt at gmail.com>:
>>> I think you're misunderstanding is_dir() and is_file(), as these don't
>>> actually call os.stat(). All DirEntry methods either call nothing or
>>> os.lstat() to get the stat info on the entry itself (not the
>>> destination of the symlink).
>>
>>
>> Oh. Extract of your PEP: "is_dir(): like os.path.isdir(), but much cheaper".
>>
>> genericpath.isdir() and genericpath.isfile() use os.stat(), whereas
>> posixpath.islink() uses os.lstat().
>>
>> Is it a mistake in the PEP?
>
> Ah, you're dead right -- this is basically a bug in the PEP, as
> DirEntry.is_dir() is not like os.path.isdir() in that it is based on
> the entry itself (like lstat), not following the link.
>
> I'll improve the wording here and update the PEP.

Ok, so it means that your example grouping files per type, files and
directories, is also wrong. Or at least, it behaves differently than
os.walk(). You should put symbolic links to directories in the "dirs"
list too.

if entry.is_dir():   # is_dir() checks os.lstat()
    dirs.append(entry)
elif entry.is_symlink() and os.path.isdir(entry):   # isdir() checks os.stat()
    dirs.append(entry)
else:
    non_dirs.append(entry)

Victor