BeanstalkError: EXPECTED_CRLF when putting UTF-8 string containing non-ASCII characters

[egorgvo]

Description

When calling:

client.put_job_into(tube_name, "café (coffee)")

I get the following traceback:

Traceback (most recent call last):
  File "/media/user/code/module.py", line 49, in push_job
    return self.put_job(  # type: ignore[no-any-return]
  File "/media/user/venv/lib/python3.10/site-packages/pystalk/client.py", line 300, in put_job
    return self._receive_id(socket)
  File "/media/user/venv/lib/python3.10/site-packages/pystalk/client.py", line 221, in _receive_id
    status, gid = self._receive_name(sock)
  File "/media/user/venv/lib/python3.10/site-packages/pystalk/client.py", line 230, in _receive_name
    raise BeanstalkError(message.rstrip())
pystalk.client.BeanstalkError: b'EXPECTED_CRLF'

This happens because the length of the string in bytes may differ from the number of characters (len(data) vs len(data.encode('utf-8'))) when the string contains non-ASCII characters like é.

---

Suggested Fix

Update put_job to encode data to bytes before calculating the length (see comment inside the code):

def put_job(self, data: Union[str, bytes], pri: int = 65536, delay: int = 0, ttr: int = 120):
    with self._sock_ctx() as socket:
        if not isinstance(data, bytes):  # <-- I just moved this check from above to here
            data = data.encode('utf-8')
        message = 'put {pri} {delay} {ttr} {datalen}\r\n'.format(
            pri=pri, delay=delay, ttr=ttr, datalen=len(data)
        ).encode('utf-8')
        message += data
        message += b'\r\n'
        self._send_message(message, socket)
        return self._receive_id(socket)

This ensures the correct byte length is used in the put command, preventing EXPECTED_CRLF errors with UTF-8 strings.

If needed, I can prepare a PR with this fix.