Why is `clean_str` present in `wikienv`?

I am finding certain strings can break `clean_str`:

```python
p = "This is a test string with unicode escape: \\u00e9"
```

This will break `clean_str`:

```none
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xe9 in position 43: unexpected end of data
```

Why do we need to convert string to be UTF-8? And if it's required, why not just `ignore` conversion errors?