Support an `encoding` parameter when reading netCDF files with xarray

Is your feature request related to a problem?

When I use netCDF4.Dataset to open a file whose path contains Chinese characters, I get:

import netCDF4
file = "文件.nc"
ds = netCDF4.Dataset(file)
# FileNotFoundError: [Errno 2] No such file or directory: '文件.nc'

If I specify encoding='gbk', it opens correctly:

ds = netCDF4.Dataset(file, encoding='gbk')

However, xarray’s open_dataset does not support an encoding argument:

import xarray as xr
file = "文件.nc"
ds = xr.open_dataset(file)
# FileNotFoundError: [Errno 2] No such file or directory: 'E:\\netcdf\\文件.nc'

ds = xr.open_dataset(file, engine='netcdf4', encoding='gbk')
# TypeError: NetCDF4BackendEntrypoint.open_dataset() got an unexpected keyword argument 'encoding'

Related netCDF4 issue comment
(from netCDF4-python#997)

Describe the solution you'd like

I would like xarray to accept an encoding parameter so that files with Chinese paths open successfully.

Describe alternatives you've considered

No response

Additional context

No response

Start at NetCDF4BackendEntrypoint.open_dataset, which currently rejects the encoding argument, and trace how xarray passes arguments to the netCDF4 backend. Done means encoding is accepted for this backend and a file path containing Chinese characters opens successfully, with coverage for the reported case.