refactor: Use pathlib on platformdirs folders

[rockstorm101]

Here is some refactoring to clean up the code base a bit. Nothing in this PR is written in stone so please feel free to reject/criticize/suggest changes. I tried to make commits self-explanatory but here is a summary:

printrun/pronsole:

• Use more modern pathlib.Path instead of os.path for platformdirs folders
• Initialize app dirs right after variable creation to avoid checking for
  their existence all the time.
• Read/write files using with statements instead of manually opening and
  closing them.
• Remove data_dir variable which is not used.

printrun/pronterface:

• History file and cache dir are created during pronsole initialization so no
  need to check for their existence.

printrun/utils:

• Update read/write history methods to use a pathlib.Path file instead of a file
  name. Existence check is removed from these functions because such check
  belongs to caller functions instead.
• Simplify lookup_file function and their respective callers
• sharedfile is not used and therefore removed

[DivingDuck]

Hi, just did a quick test on Windows and got an error:

[rockstorm101]

Hi, just did a quick test on Windows and got an error:

Many thanks for testing. I just pushed a commit hoping to fix that one. Let me know if that works now. Also fixed a regression that was picked up by the GitHub Actions.

[DivingDuck]

That one is working fine now, but found another one:

I got an logging error when Log to console is activated. I did a connect and disconnect. When I then try to close the Pronterface Window I'm not able to do so and got the second traceback w/o been able to close Pronterface. I need to kill the process manually.
Maybe also a path problem with dirs as I can't see any update in the log file.

Edit: I forgot to mention that the errors happen only if I run from the binary file - running from source is fine.

Terminal output:

Verbinde...
--- Logging error ---
Traceback (most recent call last):
Traceback (most recent call last):
Keine Verbindung zum Drucker.
--- Logging error ---
Traceback (most recent call last):
  File "logging\__init__.py", line 1153, in emit
  File "printrun\pronterface.py", line 113, in write
AttributeError: 'NoneType' object has no attribute 'write'

During handling of the above exception, another exception occurred:


Traceback (most recent call last):
  File "printrun\printcore.py", line 171, in logError
  File "printrun\pronsole.py", line 284, in logError
  File "logging\__init__.py", line 2162, in error
  File "logging\__init__.py", line 1548, in error
  File "logging\__init__.py", line 1664, in _log
  File "logging\__init__.py", line 1680, in handle
  File "logging\__init__.py", line 1736, in callHandlers
  File "logging\__init__.py", line 1026, in handle
  File "logging\__init__.py", line 1158, in emit
  File "logging\__init__.py", line 1074, in handleError
  File "printrun\pronterface.py", line 113, in write
AttributeError: 'NoneType' object has no attribute 'write'

--- Logging error ---
Traceback (most recent call last):
Traceback (most recent call last):

[rockstorm101]

I got an logging error when Log to console is activated. I did a connect and disconnect.

Can you confirm this error does not happen on the master branch as well? Because the code regarding logging is not affected by the PR I believe. I'm unable to reproduce the behavior.

[DivingDuck]

Hi,
I just check my files. It happen in the master branch too. Looking back, I'm not sure that I had test this exact combination. The error only happen when I switch from not logging to stdout to logging to stdout and then restart Pronterface and there only if I run from binaries. Running from source is fine and do not show the behavior.

I need to take a closer look what happen with introducing this change (#1473) and why it fails in such a strange way and more over why there is no entry with the error in the log file as well...

Anyway, with this observation your PR works fine and we need to address the logging problem as an new issue.

[neofelis2X]

Works also fine on macOS. 🚀

[DivingDuck]

@rockstorm101, I created a new issue for my finding regarding the logging issue #1485.