Warn when _parse_ccsds_head is given more than 10 bytes

[namurphy]

_parse_ccsds_head decodes only the first 10 bytes it is given and silently ignores the rest. This PR makes it issue a UserWarning when it gets more than 10 bytes, because a longer input probably means the caller sliced the bytes incorrectly.

read_file_sc was the only caller that passed more than 10 bytes: it passed the rest of the file (bytestr, or bytestr[17:] for PTP files) to read the first header. It now passes bytestr[:10] or bytestr[17:27], so spacecraft conversions don't raise the new warning. The other callers, _read_bytestr and _wrapper_status, already pass slices of at most 10 bytes.

Other changes:

• Adds a Warns section to the docstring of _parse_ccsds_head.
• Removes a SIM108 suppression in read_file_sc that ruff now reports as unused.
• Replaces test_parse_ccsds_head_ignores_extra_bytes with test_parse_ccsds_head_warns_about_extra_bytes, which checks that the warning is raised and that the extra bytes still don't change the result.

No test covers read_file_sc because there is no spacecraft test data, so the change to its slicing is untested. If the first header is cut short, the slice raises the same ValueError as before, which read_file_sc already catches.

🤖 Generated with Claude Code

[namurphy]

I don't think it's normal behavior for the bytearr to be longer than expected, but I may be mistaken about that. For now, issuing a warning seems appropriate, and we can re-address this later if the warnings end up being unnecessary or too noisy.