mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-07-25 17:18:11 +00:00
a23115414a
PR #55339 adds encoding='utf-8', errors='replace' to 26 subprocess.run(text=True) call sites across the codebase. The triage review (thanks @alt-glitch) diffed this PR against #55339 and found that 5 of the 6 originally-touched call sites are already covered there byte-identically: - hermes_cli/main.py::_probe_container - hermes_cli/setup.py SSH probe - tools/tts_tool.py::_generate_neutts - tools/transcription_tools.py::_prepare_local_audio - tools/transcription_tools.py::_transcribe_local_command (both branches) The one genuinely net-new site — hermes_cli/onepassword_secrets_cli.py::_op_whoami (the 1Password op CLI whoami probe) — is NOT in #55339 and is fixed here. Without explicit encoding=, text=True decodes child output with locale.getpreferredencoding(False) — cp936 on Chinese Windows — which crashes _readerthread on non-GBK bytes, cascading into pipe buffer fills, event loop stalls, and TUI freezes (issues #47939, #53428, #57238). Scope narrowed per triage feedback: the other 5 sites should land via #55339. Refs #53428 (together with #55339).
52 lines
2.2 KiB
Python
52 lines
2.2 KiB
Python
"""Regression test for issue #53428 — subprocess.run(text=True) without
|
|
explicit ``encoding=`` triggers ``UnicodeDecodeError`` on Chinese Windows
|
|
(cp936/GBK default encoding).
|
|
|
|
PR #55339 covers 21 call sites in ``agent/``, ``gateway/``, ``cli.py``,
|
|
``cron/``, plus 5 more in ``tools/`` and ``hermes_cli/`` (main.py,
|
|
setup.py, tts_tool.py, transcription_tools.py). The one call site it
|
|
misses — ``hermes_cli/onepassword_secrets_cli.py::_op_whoami`` — is
|
|
guarded here.
|
|
|
|
Without ``encoding=``, ``text=True`` decodes child output with
|
|
``locale.getpreferredencoding(False)`` — cp936 on Chinese Windows —
|
|
which crashes on non-GBK bytes (issues #47939, #53428, #57238).
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
from pathlib import Path
|
|
from unittest.mock import patch, MagicMock
|
|
|
|
|
|
def _assert_utf8_kwargs(mock_run):
|
|
"""Lift the encoding/errors kwargs out of the mock and assert them."""
|
|
assert mock_run.called, "subprocess.run was not called"
|
|
kwargs = mock_run.call_args.kwargs
|
|
assert kwargs.get("encoding") == "utf-8", (
|
|
f"subprocess.run called without encoding='utf-8' "
|
|
f"(got encoding={kwargs.get('encoding')!r}). "
|
|
f"On Chinese Windows (cp936), text=True without explicit encoding "
|
|
f"crashes with UnicodeDecodeError on non-GBK bytes. See #53428."
|
|
)
|
|
assert kwargs.get("errors") == "replace", (
|
|
f"subprocess.run called without errors='replace' "
|
|
f"(got errors={kwargs.get('errors')!r}). encoding='utf-8' alone "
|
|
f"still raises UnicodeDecodeError on non-UTF-8 bytes emitted by "
|
|
f"Windows-native CLIs. See #53428."
|
|
)
|
|
|
|
|
|
def test_op_whoami_passes_utf8_encoding(tmp_path):
|
|
"""_op_whoami must pass encoding='utf-8', errors='replace' so op CLI
|
|
output containing non-ASCII account names doesn't crash on cp936."""
|
|
from hermes_cli import onepassword_secrets_cli as op_cli
|
|
|
|
fake_binary = tmp_path / "op"
|
|
fake_binary.write_bytes(b"")
|
|
with patch.object(op_cli.subprocess, "run") as mock_run:
|
|
mock_run.return_value = MagicMock(
|
|
returncode=0, stdout="user@example.com", stderr=""
|
|
)
|
|
op_cli._op_whoami(fake_binary, account="")
|
|
_assert_utf8_kwargs(mock_run)
|