mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-07-31 19:16:29 +00:00
Second, deeper pass over tools/gateway/hermes_cli plus first pass over the trees wave 1 missed (acp, acp_adapter, skills, computer_use, docker, dashboard, conformance, monitoring, secret_sources, hermes_state, providers). Same rubric as wave 1 (AGENTS.md test policy); security, alternation/caching invariants, issue-number regressions, and E2E kept. Real test-quality fixes found and rooted out along the way: - tests/tools/test_command_guards.py made real auxiliary-LLM HTTPS calls (DEFAULT_CONFIG smart-approval leaked in) — pinned approval mode=manual via autouse fixture: 17.4s → 0.4s. - test_model_switch_custom_providers.py / test_user_providers_model_switch.py silently probed live provider catalogs (~2s/test) — stubbed cached_provider_model_ids/provider_model_ids/fetch_api_models. - test_telegram_noise_filter.py: 15-platform copy-paste matrix over shared gateway.run logic → 3 representative platforms (55s → 3.9s). - test_gateway_shutdown.py: stop()'s 5s interrupt-deadline loop spun on MagicMock agents — interrupt.side_effect now clears _running_agents (22s → 1.0s). - test_gateway_inactivity_timeout.py poll-harness timings shrunk 3-5x (24s → 1.1s); test_mcp_stability.py backoff/SIGTERM-grace sleeps patched (15.4s → 2.5s); test_async_delegation.py negative-drain wait 5s → 0.5s. - test_telegram_init_deadline.py: loop-block margin restored to 1.0s with rationale comment — the watchdog-dump assertion needs the loop blocked well past deadline+grace under parallel load (flaked once in the 40-worker verification run at a 0.2s margin). Verification: full hermetic suite via scripts/run_tests.sh — 2,438 files, 21,718 tests passed, 0 failed, 293.9s wall. Suite totals vs original baseline: 46,820 → 19,757 test functions (−57.8%), wall 583.5s → 293.9s (−50%), subprocess CPU 13,564s → 11,623s.
177 lines
5.7 KiB
Python
177 lines
5.7 KiB
Python
"""Tests for CRLF line-ending preservation in write_file and patch.
|
|
|
|
Without this, the agent silently normalizes Windows-line-ending files
|
|
to LF whenever it edits them — and patch produces a mixed-ending file
|
|
when only a substituted region changes (the rest of the file keeps its
|
|
CRLF endings while the replacement is LF-only).
|
|
|
|
See issue #507 (Roo Code deep-dive, item 2c).
|
|
"""
|
|
|
|
import json
|
|
|
|
import pytest
|
|
|
|
|
|
@pytest.fixture
|
|
def hermes_home(monkeypatch, tmp_path):
|
|
"""Isolate HERMES_HOME so the tests don't pollute the real config.
|
|
|
|
Also clears module-level caches (file_ops, active_environments,
|
|
file-staleness state) after the test so subsequent tests in the
|
|
same pytest process aren't affected by our shell-out side effects
|
|
(real file_ops and terminal environments get created under
|
|
task_id='default' via _resolve_container_task_id).
|
|
"""
|
|
home = tmp_path / "hermes"
|
|
home.mkdir()
|
|
monkeypatch.setenv("HERMES_HOME", str(home))
|
|
yield home
|
|
# Cleanup: drop the cached file_ops and active environment so the
|
|
# next test sees a fresh state. Without this, _get_live_tracking_cwd
|
|
# returns the stale cwd from this test's ops and breaks tests like
|
|
# test_resolve_path that rely on TERMINAL_CWD env var.
|
|
try:
|
|
from tools.file_tools import clear_file_ops_cache, _read_tracker_lock, _read_tracker
|
|
clear_file_ops_cache()
|
|
with _read_tracker_lock:
|
|
_read_tracker.clear()
|
|
except Exception:
|
|
pass
|
|
try:
|
|
from tools.terminal_tool import _active_environments, _env_lock
|
|
with _env_lock:
|
|
_active_environments.clear()
|
|
except Exception:
|
|
pass
|
|
|
|
|
|
def _crlf_count(b: bytes) -> int:
|
|
return b.count(b"\r\n")
|
|
|
|
|
|
def _bare_lf_count(b: bytes) -> int:
|
|
return b.count(b"\n") - b.count(b"\r\n")
|
|
|
|
|
|
class TestPatchCRLFPreservation:
|
|
def test_patch_on_crlf_file_stays_pure_crlf(self, hermes_home, tmp_path):
|
|
"""LLM sends LF old/new; file has CRLF. Result must be all CRLF,
|
|
no mixed endings."""
|
|
from tools.file_tools import _handle_patch
|
|
|
|
target = tmp_path / "config.ini"
|
|
target.write_bytes(b"[a]\r\nkey=1\r\n\r\n[b]\r\nkey=2\r\n")
|
|
|
|
result = _handle_patch(
|
|
{
|
|
"mode": "replace",
|
|
"path": str(target),
|
|
"old_string": "key=1",
|
|
"new_string": "key=99",
|
|
},
|
|
task_id="crlf_patch_1",
|
|
)
|
|
d = json.loads(result)
|
|
assert not d.get("error"), d
|
|
|
|
raw = target.read_bytes()
|
|
assert _bare_lf_count(raw) == 0, (
|
|
f"Mixed line endings after patch: {raw!r}"
|
|
)
|
|
# Same number of line breaks as before; just the value swapped.
|
|
assert _crlf_count(raw) == 5
|
|
assert b"key=99\r\n" in raw
|
|
|
|
|
|
def test_patch_multiline_replacement_on_crlf(self, hermes_home, tmp_path):
|
|
"""Multi-line new_string with bare LFs should be CRLF-converted
|
|
before write."""
|
|
from tools.file_tools import _handle_patch
|
|
|
|
target = tmp_path / "f.py"
|
|
target.write_bytes(b"def foo():\r\n return 1\r\n")
|
|
|
|
result = _handle_patch(
|
|
{
|
|
"mode": "replace",
|
|
"path": str(target),
|
|
"old_string": "def foo():\n return 1",
|
|
"new_string": "def foo():\n x = 1\n return x",
|
|
},
|
|
task_id="crlf_patch_3",
|
|
)
|
|
d = json.loads(result)
|
|
assert not d.get("error"), d
|
|
|
|
raw = target.read_bytes()
|
|
assert _bare_lf_count(raw) == 0, (
|
|
f"Mixed endings after multi-line patch: {raw!r}"
|
|
)
|
|
assert raw == b"def foo():\r\n x = 1\r\n return x\r\n"
|
|
|
|
|
|
class TestWriteFileCRLFPreservation:
|
|
def test_overwrite_crlf_file_with_lf_content_preserves_crlf(
|
|
self, hermes_home, tmp_path
|
|
):
|
|
"""The agent typically sends bare-LF content; if the file existed
|
|
with CRLF, the write should convert to CRLF rather than silently
|
|
flipping the endings."""
|
|
from tools.file_tools import _handle_write_file
|
|
|
|
target = tmp_path / "config.bat"
|
|
target.write_bytes(b"@echo off\r\nset X=1\r\n")
|
|
|
|
result = _handle_write_file(
|
|
{
|
|
"path": str(target),
|
|
"content": "@echo off\nset X=99\nset Y=42\n",
|
|
},
|
|
task_id="crlf_write_1",
|
|
)
|
|
d = json.loads(result)
|
|
assert "error" not in d, d
|
|
|
|
raw = target.read_bytes()
|
|
assert _bare_lf_count(raw) == 0, (
|
|
f"CRLF file got normalized to LF: {raw!r}"
|
|
)
|
|
assert _crlf_count(raw) == 3
|
|
|
|
|
|
def test_overwrite_lf_file_stays_lf(self, hermes_home, tmp_path):
|
|
"""Pre-existing LF file should not get spurious CRLFs."""
|
|
from tools.file_tools import _handle_write_file
|
|
|
|
target = tmp_path / "lf.txt"
|
|
target.write_bytes(b"line1\nline2\n")
|
|
|
|
result = _handle_write_file(
|
|
{"path": str(target), "content": "X\nY\nZ\n"},
|
|
task_id="crlf_write_3",
|
|
)
|
|
d = json.loads(result)
|
|
assert "error" not in d, d
|
|
|
|
raw = target.read_bytes()
|
|
assert _crlf_count(raw) == 0
|
|
assert raw == b"X\nY\nZ\n"
|
|
|
|
|
|
class TestLineEndingHelpers:
|
|
"""Direct unit tests for the pure helpers — easier to debug than the
|
|
integration tests above."""
|
|
|
|
def test_detect_crlf(self):
|
|
from tools.file_operations import _detect_line_ending
|
|
|
|
assert _detect_line_ending("a\r\nb\r\n") == "\r\n"
|
|
|
|
|
|
def test_normalize_to_crlf_idempotent(self):
|
|
from tools.file_operations import _normalize_line_endings
|
|
|
|
once = _normalize_line_endings("a\nb\n", "\r\n")
|
|
twice = _normalize_line_endings(once, "\r\n")
|
|
assert once == twice == "a\r\nb\r\n"
|