UnicodeDecodeError on files in archive if filename contains non-ASCII characters (e.g. CJK characters)

In `libarchive/adapters/archive_read.py` line 61 the codec is set to `_ASCII_ENCODING`, which in line 18 is hard-coded into 'ascii':

```python
filepath = filepath.encode(_ASCII_ENCODING)
```
```python
_ASCII_ENCODING = 'ascii'
```

Replacing `_ASCII_ENCODING` with 'utf-8' temporarily solves my problem.