UnicodeDecodeError: 'utf-8' codec can't decode byte (Windows CHCP check)

[InfoSCE]

Hi,

I recently discover your third-party for rclone and i really enjoy it. I just want to share with you my experience when i tried to run for the first time rclonesync on my server. I read your documentation and configured properly my rclone's remote in version 1.55 and installed Python 3.9.6. I've seen i needed in Windows environnement to set "chcp to 65001" and "set PYTHONIONECODING=UTF-8" before i can run my first sync. I did it. But, when i launched my command

rclonesync -1 C:\xxx\ remote:folder

I got this error :

File "E:\RClone\v1.55_x64\rclonesync.py", line 801, in <module>
chcp = subprocess.check_output(["chcp"], shell=True).decode("utf-8")
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xff in position 20: invalid start byte

I've tried to understand why, so i checked my chcp and PYTHONIOENCODING values, all good. I opened rclonesync file with Python IDLE and i gone to line 801 :

chcp = subprocess.check_output(["chcp"], shell=True).decode("utf-8")

I've tried to run that line in python's console, same error. So, I ran the first part before "decode("utf-8)" and i got that :

b'Page de codes active\xff: 65001\r\n'

You can see the xff part which seems to generate an error. To avoid this i added an ignore argument in the command line and it works perfectly.

chcp = subprocess.check_output(["chcp"], shell=True).decode("utf-8","ignore")

But the error is still there, so if you have solution for me, i take it with pleasure :)

Thank you... and sorry for my english ;)

Damien

[cjnaz]

Thanks for the bug report, @InfoSCE.

Since my system uses English, I seldom run into utf-8 encoding, and it is difficult for me to do any debug and verification. I don't understand how to control the Windows command response so that it only returns utf-8 encoding versus codes such as 0xff. Your solution looks good for this specific error.

Note that there may be cases where you have file names that contain non-utf-8 characters and rclonesync may error out on these as well.

I will leave this issue open for others to find, if needed. Note that there is a beta version of rclone that incorporates rclonesync directly as bisync. This new rclone will replace rclonesync hopefully later this year. As such, I'm doing only critical bug fixes on rclonesync.

[cjnaz]

This error (fixed with the .decode("utf-8","ignore")) would come up in all runs, not just the first sync. I've updated the issue title.