mirror of
https://github.com/NousResearch/hermes-agent.git
synced 2026-07-31 19:16:29 +00:00
6b81590c55
Systematic prune per AGENTS.md test policy, one pass over every major test tree (gateway, hermes_cli, tools, agent, run_agent, plugins, cli, cron, tui_gateway, honcho/openviking, root-level): - DELETE: source-reading tests (read_text/getsource on prod files), change-detector tests (exact catalog counts, model-name snapshots, config version literals), mock-echo tests (assert a mock returns what it was told), assertion-free/trivial tests, near-duplicate parametrizations (boundaries + one representative kept), async/sync twin duplicates, cosmetic within-file variations. - KEEP (mandatory): security/redaction/approval guards, message-role alternation invariants, prompt-caching/deterministic-call-id invariants, issue-number regression tests (deduped), E2E tests. - 6 test files deleted outright (script-style/no-assert or fully redundant); conftest.py, fakes/, fixtures/ untouched. - tests/acp/conftest.py added: autouse fixture stubs the live models.dev/GitHub/Copilot/Anthropic inventory fetches that ACP server tests performed on every session create — test_server.py 147s → 3.4s, and the tests are now genuinely hermetic. - Sleep-based slowness shrunk where safe (codex_ttfb_watchdog, compression_concurrent_fork, etc.); no wall-clock assertion tightened. Verification: full hermetic suite via scripts/run_tests.sh — 2439 files, 31,130 tests passed, 0 failed, 0 flaky retries, 315s wall (baseline: 583s wall, 13,564s subprocess CPU).
105 lines
3.3 KiB
Python
105 lines
3.3 KiB
Python
"""Tests for /undo handling in tui_gateway.
|
|
|
|
The TUI routes ``/undo`` through ``command.dispatch`` (it's in
|
|
``_PENDING_INPUT_COMMANDS`` because the CLI handler queues input the
|
|
slash-worker subprocess can't read). The server handles it directly,
|
|
mutates SessionDB to soft-delete rows, refreshes the in-memory session
|
|
history, fires the memory-provider hook with ``rewound=True``, and
|
|
returns ``{"type": "prefill", "message": <text>, "notice": ...}`` so
|
|
the Ink client drops the message into the composer for editing.
|
|
|
|
``/undo N`` backs up N user turns at once (default 1). See issue #21910.
|
|
"""
|
|
|
|
from __future__ import annotations
|
|
|
|
import importlib
|
|
import threading
|
|
from pathlib import Path
|
|
from unittest.mock import MagicMock, patch
|
|
|
|
import pytest
|
|
|
|
from hermes_state import SessionDB
|
|
|
|
|
|
@pytest.fixture()
|
|
def hermes_home(tmp_path, monkeypatch):
|
|
home = tmp_path / ".hermes"
|
|
home.mkdir()
|
|
monkeypatch.setattr(Path, "home", lambda: tmp_path)
|
|
monkeypatch.setenv("HERMES_HOME", str(home))
|
|
yield home
|
|
|
|
|
|
@pytest.fixture()
|
|
def server(hermes_home):
|
|
with patch.dict(
|
|
"sys.modules",
|
|
{
|
|
"hermes_cli.env_loader": MagicMock(),
|
|
"hermes_cli.banner": MagicMock(),
|
|
},
|
|
):
|
|
mod = importlib.import_module("tui_gateway.server")
|
|
yield mod
|
|
# Reset module-level session state without re-importing. importlib.reload
|
|
# would re-register the module's atexit hooks; duplicated hooks race the
|
|
# stderr buffer at interpreter shutdown (Fatal Python error:
|
|
# _enter_buffered_busy) — same class as PR #34217.
|
|
mod._sessions.clear()
|
|
mod._pending.clear()
|
|
mod._answers.clear()
|
|
# NOTE: _methods is intentionally NOT cleared — it's populated at import
|
|
# time and would only repopulate via reload.
|
|
mod._db = None
|
|
|
|
|
|
@pytest.fixture()
|
|
def db(hermes_home):
|
|
return SessionDB(db_path=hermes_home / "state.db")
|
|
|
|
|
|
@pytest.fixture()
|
|
def session_with_history(server, db):
|
|
"""Build a session with 3 user turns + assistant replies persisted in DB."""
|
|
sid = "sid-undo"
|
|
session_key = "tui-undo-1"
|
|
db.create_session(session_key, source="tui")
|
|
for i in range(1, 4):
|
|
db.append_message(session_key, "user", f"question {i}")
|
|
db.append_message(session_key, "assistant", f"answer {i}")
|
|
history = db.get_messages_as_conversation(session_key)
|
|
agent = MagicMock()
|
|
agent._memory_manager = MagicMock()
|
|
agent._last_flushed_db_idx = len(history)
|
|
s = {
|
|
"session_key": session_key,
|
|
"history": list(history),
|
|
"history_lock": threading.Lock(),
|
|
"history_version": 0,
|
|
"running": False,
|
|
"agent": agent,
|
|
"attached_images": [],
|
|
"cols": 120,
|
|
}
|
|
server._sessions[sid] = s
|
|
# Wire the DB cache so _get_db() returns our fixture.
|
|
server._db = db
|
|
return sid, session_key, s, agent
|
|
|
|
|
|
def _call(server, method, **params):
|
|
return server._methods[method](1, params)
|
|
|
|
|
|
def test_undo_returns_prefill_with_target_text(server, session_with_history):
|
|
sid, session_key, s, agent = session_with_history
|
|
resp = _call(server, "command.dispatch", session_id=sid, name="undo", arg="")
|
|
result = resp["result"]
|
|
assert result["type"] == "prefill"
|
|
# Default /undo backs up one user turn — "question 3"
|
|
assert result["message"] == "question 3"
|
|
assert "Undid" in result["notice"]
|
|
|
|
|