BYTES directive leads to exception when the bytes aren't UTF-8 decodeable

[BT8298]

An exception is thrown anytime a string appears in a BYTES directive (#220) that is not utf-8 decodeable. Using the dialogue pair

- q: "BYTES(\xaa\x55\x00\x02\x08\x01\xf4)"
  r: "BYTES(\xaa\x55\x00\x02\x08\x01\xf4)"

my project's test suite fails with the traceback

Traceback (most recent call last):
  File "/home/user/python-project/tests/test_tpi.py", line 20, in setUp
    self.tpi = package.aux_control.TPI_RF_Generator("ASRL1::INSTR")
               ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^
  File "/home/user/python-project/package/aux_control.py", line 876, in __init__
    if self._query(TPI_CMD["set_user_control"]) != TPI_CMD["set_user_control"]:
       ~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/python-project/package/aux_control.py", line 1093, in _query
    self._write(command, value)
    ~~~~~~~~~~~^^^^^^^^^^^^^^^^
  File "/home/user/python-project/package/aux_control.py", line 919, in _write
    bytes_written = self._instrument.write_raw(packet)
  File "/home/user/python-project/.venv/lib/python3.14/site-packages/pyvisa/resources/messagebase
d.py", line 173, in write_raw
    return self.visalib.write(self.session, message)[0]
           ~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/user/python-project/.venv/lib/python3.14/site-packages/pyvisa_sim/highlevel.py", li
ne 290, in write
    return sess.write(data)  # type: ignore
           ~~~~~~~~~~^^^^^^
  File "/home/user/python-project/.venv/lib/python3.14/site-packages/pyvisa_sim/sessions/serial.p
y", line 59, in write
    self.device.write(val)
    ~~~~~~~~~~~~~~~~~^^^^^
  File "/home/user/python-project/.venv/lib/python3.14/site-packages/pyvisa_sim/devices.py", line
 317, in write
    response = self._match(message)
  File "/home/user/python-project/.venv/lib/python3.14/site-packages/pyvisa_sim/devices.py", line
 423, in _match
    response = self._match_setters(query)
  File "/home/user/python-project/.venv/lib/python3.14/site-packages/pyvisa_sim/component.py", li
ne 418, in _match_setters
    q = query.decode("utf-8")
UnicodeDecodeError: 'utf-8' codec can't decode byte 0xaa in position 0: invalid start byte

I was able to solve this by replacing a few instances of utf-8 codec in the source with latin-1. However, this limits the dialogue definitions and properties to the latin-1 codec only, preventing simulation of devices that use characters outside of latin-1. This is why I open the issue: to ask whether this change could negatively impact some users. If anyone knows of devices relevant to PyVISA which use utf-8 strings in the I/O protocol, please mention it in a reply. Also, feedback on alternate implementation of the fix are welcome.

[MatthieuDartiailh]

One option could be to always match on encoded bytes rather than decoded strings. And then using utf8 to encode non byte string in the yaml definition should be safe.

[BT8298]

That approach could be challenging due to reliance on stringparser library. For dialogues it seems fine as those are handled as bytes (except when there is a RANDOM directive), but matches to setters are detected by passing the decoded query string to a stringparser.Parser. I don't think stringparser has support for bytes.