The issue isn’t the character, itself, since we fully support and have tested these types of characters in our SFTP integration.
The issue is that there are filenames that are not encoded as expected. They may be ISO-8859-1 vs UTF-8, for example. This is a bit odd since default filename encoding is usually going to be UTF-8/ASCII, especially on Linux systems, but there may be something that is writing files to that system in a different encoding, somehow.
In order for the odrive SFTP integration to work properly, UTF-8 (or a subset of it) are needed.
I’ve been looking at ways to check for the filename encoding, and this method seems to work, mostly. It is an ugly one-line python script you can run on the server in a given folder. If you copy and paste this command it will go through each item in the current directory and spit out the “estimated” filename encoding.
python -c "(lambda __g, __y, __print: [[(lambda __after, __items, __sentinel: __y(lambda __this: lambda: (lambda __i: [(__print(('%s => %s (%s)' % (n, chardet.detect(n)['encoding'], chardet.detect(n)['confidence']))), __this()) for __g['n'] in [(__i)]] if __i is not __sentinel else __after())(next(__items, __sentinel)))())(lambda: None, iter(os.listdir('.')), ) for __g['os'] in [(__import__('os', __g, __g))]] for __g['chardet'] in [(__import__('chardet', __g, __g))]])(globals(), (lambda f: (lambda x: x(x))(lambda y: f(lambda: y(y)()))), __import__('__builtin__', level=0).__dict__['print'])"
The diagnostic shows issues listing the root folder of your SFTP link and the “P Docs” folder, but there may be others. If you run the above command in those folders, do you see any encodings that are not ascii or utf-8?