:flat 1 crashes ranger

[vejkse]

Runtime Environment

• Operating system and version: Linux 6.2.7
• Terminal emulator and version: Alacritty
• Python version: 3.10.10
• Ranger version/commit: 8875ad2 (the problem arises already with 46660c2)
• Locale: mixture of en_US.UTF-8, en_DK.UTF-8, POSIX

Current Behavior

When executing :flat 1 in any directory, ranger crashes. Tested with ranger --clean.

Expected Behavior

Not crashing and displaying all files up to depth 1.

Possible Solutions

Rollback to a previous version. I have commit f8b304f available and it works fine.

Traceback

ranger version: ranger-master
Python version: 3.10.10 (main, Mar  5 2023, 22:26:53) [GCC 12.2.1 20230201]
Locale: en_US.UTF-8
Current file: '/home'

Traceback (most recent call last):
  File "/usr/lib/python3.10/site-packages/ranger/core/main.py", line 204, in main
    fm.loop()
  File "/usr/lib/python3.10/site-packages/ranger/core/fm.py", line 417, in loop
    ui.redraw()
  File "/usr/lib/python3.10/site-packages/ranger/gui/ui.py", line 343, in redraw
    self.draw()
  File "/usr/lib/python3.10/site-packages/ranger/gui/ui.py", line 370, in draw
    DisplayableContainer.draw(self)
  File "/usr/lib/python3.10/site-packages/ranger/gui/displayable.py", line 268, in draw
    displayable.draw()
  File "/usr/lib/python3.10/site-packages/ranger/gui/widgets/view_miller.py", line 98, in draw
    directory.load_content_if_outdated()
  File "/usr/lib/python3.10/site-packages/ranger/container/directory.py", line 678, in load_content_if_outdated
    real_mtime = mtimelevel(self.path, self.flat)
  File "/usr/lib/python3.10/site-packages/ranger/container/directory.py", line 89, in mtimelevel
    for dirpath, dirnames, _ in walklevel(path, level):
  File "/usr/lib/python3.10/site-packages/ranger/container/directory.py", line 78, in walklevel
    assert os.path.isdir(some_dir)
AssertionError

[vejkse]

I sometimes have a slightly different error:

Traceback (most recent call last):
  File "/usr/lib/python3.10/site-packages/ranger/core/main.py", line 204, in main
    fm.loop()
  File "/usr/lib/python3.10/site-packages/ranger/core/fm.py", line 417, in loop
    ui.redraw()
  File "/usr/lib/python3.10/site-packages/ranger/gui/ui.py", line 343, in redraw
    self.draw()
  File "/usr/lib/python3.10/site-packages/ranger/gui/ui.py", line 370, in draw
    DisplayableContainer.draw(self)
  File "/usr/lib/python3.10/site-packages/ranger/gui/displayable.py", line 268, in draw
    displayable.draw()
  File "/usr/lib/python3.10/site-packages/ranger/gui/widgets/view_miller.py", line 98, in draw
    directory.load_content_if_outdated()
  File "/usr/lib/python3.10/site-packages/ranger/container/directory.py", line 678, in load_content_if_outdated
    real_mtime = mtimelevel(self.path, self.flat)
  File "/usr/lib/python3.10/site-packages/ranger/container/directory.py", line 92, in mtimelevel
    mtime = max(mtime, [-1] + [os.stat(d).st_mtime for d in dirlist])
TypeError: '>' not supported between instances of 'list' and 'float'

Combined with the fact that the only recent change in the file directory.py is this one line: 49fc14d, it seems clear that the removed max was necessary.

I changed back by hand this line and everything works fine.

[vejkse]

I think I know where this comes from: I just ran pylint directory.py and one of the messages is

directory.py:92:16: W3301: Do not use nested call of 'max'; it's possible to do 'max(mtime, [-1] + [os.stat(d).st_mtime for d in dirlist])' instead (nested-min-max)

Clearly it shouldn’t suggest that. There is in fact a recent pylint bug about that: pylint-dev/pylint#8168.

[vejkse]

So, after looking at the pylint bug, which is not solved for this specific case, rather than reverting the commit 49fc14d, it suffices to add a * in the line:

mtime = max(mtime, *[-1] + [os.stat(d).st_mtime for d in dirlist])

(or perhaps simply to move mtime, inside the square bracket).

[toonn]

Thank you for bringing this to our attention!

I agree we should just append mtime instead of the -1.

[toonn]

I've opened a PR, tried to add you as a co-author on the commit with a GitHub no-reply email. Could you confirm I used the correct committer email address for you, or whether you'd prefer not to be listed in the first place?

[vejkse]

Sorry, I just saw this, having been busy these last few days. I think I’d just prefer not to be listed.