steam-backlog-enforcer/steam_backlog_enforcer/tests/conftest.py
Krzysztof kuhy Rudnicki b3e629e3f0 Bill gaming time only while the user is actually playing
A running game process was treated as proof of play. On 2026-08-28 the budget
billed ~2h23m while a gatelock screen locker covered the display and the
machine could not be touched at all -- the game was simply still resident.

A tick now bills only when all three hold, evaluated together rather than
short-circuited so the audit log can answer "was the screen ALSO locked?":

  * no gatelock guard owns the screen  (_screen_hold, via /proc/locks)
  * the focused window belongs to a qualifying process or its ancestry
  * some input source has been used within idle_grace_seconds (default 300)

Any probe that cannot answer BILLS the tick and is recorded in `degraded`.
A broken detector must never become an unlimited-gaming exploit.

Three findings from measuring this machine rather than assuming:

  * /dev/input/js0 here is an *ASRock LED Controller* -- the kernel exposes
    the motherboard RGB device as a joystick. A presence check would have
    disabled the idle criterion permanently, so _controller_input watches for
    real events instead (skipping joydev's synthetic open burst).
  * Aseprite's window carries no _NET_WM_PID at all, so focus attribution
    falls back to XRes, which asks the server who owns the window rather than
    trusting the client to have announced itself.
  * /proc/locks reports gatelock's hold without taking it; gatelock's own
    read_claim_if_held detects held-ness by acquiring, which would race a
    guard every 3 seconds.

Idle pauses also refund the grace period once per walk-away, on the
engaged->paused edge only, never across a day rollover, and never on the first
tick after a restart -- the only place the counter moves backwards.

Everything is recorded to /var/log/steam-backlog-enforcer/budget.jsonl, on
change plus a heartbeat rather than one line per 3s tick. Records name the
qualifying PIDs: reconstructing this incident from journald alone was only
possible by inferring accrual from the spacing of two warning notifications.

Also: refuse to bounce Steam while a game is running. ensure_steam_debug_port
relaunches Steam when the CDP port is closed, which kills the running game --
that is how restarting the daemon to deploy this change ended a live session.
It now defers exactly as it already did for in-flight game updates.

Talks to X through ctypes over one cached connection: at a 3s tick, forking
xprintidle and xprop would cost ~40 forks a minute.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019CEsG6BzHtYR9cHcdxPZqG
2026-08-28 19:24:14 +02:00

225 lines
9.1 KiB
Python

"""Safety conftest: prevent tests from touching real Steam/config files.
Redirects all filesystem paths used by the steam_backlog_enforcer package
to temporary directories. This stops tests from accidentally:
- Deleting real game files via uninstall_other_games / uninstall_game
- Overwriting ~/.config/steam_backlog_enforcer/state.json (losing the
user's current assignment)
- Reading real appmanifest files from ~/.local/share/Steam/steamapps
- Modifying /etc/hosts via the store blocker
- Corrupting the HLTB cache on disk
- Launching real Steam or calling real subprocess commands
- Deleting real ~/.steam, ~/.local/share/Steam, etc. via the total-block
Steam/Proton remnant purge
"""
from __future__ import annotations
from typing import TYPE_CHECKING
from unittest.mock import MagicMock, patch
import pytest
# Imported for its autouse side effect: naming it here registers it.
from steam_backlog_enforcer.tests._isolate_playtime import _isolate_playtime
from steam_backlog_enforcer.tests._no_subprocess import _block_real_subprocesses
# Re-exported so ruff --fix does not delete the imports above: pytest
# registers autouse fixtures by name, so they look unused to the linter.
__all__ = ["_block_real_subprocesses", "_isolate_playtime"]
from steam_backlog_enforcer.tests._total_block_paths import (
BLOCK,
HOSTS,
IPTABLES,
PURGE,
Paths,
build_paths,
)
if TYPE_CHECKING:
from collections.abc import Iterator
from pathlib import Path
@pytest.fixture(autouse=True)
def _isolate_filesystem(tmp_path: Path) -> Iterator[None]:
"""Redirect all real filesystem paths to a temporary directory.
Individual tests that also patch these paths will simply override
this fixture's patches for the duration of their own ``with`` block.
"""
fake_config = tmp_path / "config"
fake_config.mkdir()
fake_steamapps = tmp_path / "steamapps"
fake_steamapps.mkdir()
fake_hosts = tmp_path / "hosts"
with (
# Config / state / snapshot paths (used by State.save, Config.save, etc.)
patch(
"steam_backlog_enforcer.config.CONFIG_DIR",
fake_config,
),
patch(
"steam_backlog_enforcer.config.CONFIG_FILE",
fake_config / "config.json",
),
patch(
"steam_backlog_enforcer.config.STATE_FILE",
fake_config / "state.json",
),
patch(
"steam_backlog_enforcer.config.SNAPSHOT_FILE",
fake_config / "snapshot.json",
),
# Steam game manifests / install dirs. STEAMAPPS_PATH is canonical in
# _steam_state; game_install imports it, so both bindings need the
# patch (same gotcha as HOSTS_FILE below).
patch(
"steam_backlog_enforcer._steam_state.STEAMAPPS_PATH",
fake_steamapps,
),
patch(
"steam_backlog_enforcer.game_install.STEAMAPPS_PATH",
fake_steamapps,
),
# HLTB cache file (computed at import time from CONFIG_DIR, so
# patching CONFIG_DIR alone does not redirect it)
patch(
"steam_backlog_enforcer._hltb_types.HLTB_CACHE_FILE",
fake_config / "hltb_cache.json",
),
# /etc/hosts (store blocker + total block - each module has its own
# `from ... import HOSTS_FILE` binding, so each needs its own patch)
patch(
"steam_backlog_enforcer.store_blocker.HOSTS_FILE",
fake_hosts,
),
patch(
"steam_backlog_enforcer._total_block_hosts.HOSTS_FILE",
fake_hosts,
),
patch(
"steam_backlog_enforcer.config.HOSTS_FILE",
fake_hosts,
),
# Total-block lock + IP cache (computed at import time from
# CONFIG_DIR, so patching CONFIG_DIR alone does not redirect them -
# same gotcha as HLTB_CACHE_FILE above. A real total-block lock may
# be active on the host machine; tests must never touch it.)
patch(
"steam_backlog_enforcer._total_block.TOTAL_BLOCK_LOCK_FILE",
fake_config / "total_block_lock.json",
),
patch(
"steam_backlog_enforcer._total_block_iptables._IPTABLES_IP_CACHE_FILE",
fake_config / "total_block_ip_cache.json",
),
patch(
"steam_backlog_enforcer._total_block_purge._STEAM_PURGE_LOG_FILE",
fake_config / "total_block_purge_log.json",
),
# Steam/Proton remnant paths (real ~/.steam, ~/.local/share/Steam,
# etc. - _remove_steam_remnants() deletes these, so tests must never
# see the real ones)
patch(
"steam_backlog_enforcer._total_block_purge._STEAM_REMNANT_PATHS",
(
tmp_path / "fake_home" / ".steam",
tmp_path / "fake_home" / "steam",
tmp_path / "fake_home" / ".local" / "share" / "Steam",
tmp_path / "fake_home" / ".steampath",
tmp_path / "fake_home" / ".steampid",
tmp_path / "fake_home" / ".config" / "steamtinkerlaunch",
tmp_path / "fake_home" / ".config" / "CSDSteamBuild",
),
),
# Derived from CONFIG_DIR at import time, so it captures the REAL path
# before this fixture redirects CONFIG_DIR. Without it the suite
# rewrites the user's owned-games cache.
patch(
"steam_backlog_enforcer._owned_apps_cache._OWNED_IDS_CACHE_FILE",
fake_config / "owned_app_ids_cache.json",
),
# Whitelist exception files (module-level constants; these live in
# _whitelist_locking, which is where the code that reads them is)
patch(
"steam_backlog_enforcer._whitelist_locking.APPROVED_EXCEPTIONS_FILE",
fake_config / "approved_exceptions.json",
),
patch(
"steam_backlog_enforcer._whitelist_locking.EXCEPTION_AUDIT_LOG",
fake_config / "exception_audit.log",
),
# _enforce_loop imports CONFIG_FILE directly; patch the local binding so
# lock_enforcement_files() uses the tmp path instead of the real one.
patch(
"steam_backlog_enforcer._enforce_loop.CONFIG_FILE",
fake_config / "config.json",
),
):
yield
@pytest.fixture(autouse=True)
def _no_real_sleep() -> Iterator[None]:
"""No-op every ``time.sleep`` used by the package.
Several modules call ``time.sleep`` for Steam-launch / install-retry /
rate-limit pacing. Individual tests that need to observe sleep
behaviour can override these patches inside their own ``with`` block.
"""
noop = MagicMock()
with (
patch("steam_backlog_enforcer._steam_client.time.sleep", noop),
patch("steam_backlog_enforcer._steam_launch.time.sleep", noop),
patch("steam_backlog_enforcer._steam_api_client.time.sleep", noop),
patch("steam_backlog_enforcer._enforce_loop.time.sleep", noop),
):
yield
# ──────────────────────────────────────────────────────────────
# Total-block path redirection
#
# Lives here rather than in a helper module because pytest resolves
# fixtures by name: an imported fixture looks unused to ruff, and this
# repo runs `ruff --fix --unsafe-fixes`, which deletes the import.
# ──────────────────────────────────────────────────────────────
@pytest.fixture(autouse=True)
def total_block_paths(tmp_path: Path) -> Iterator[Paths]:
"""Redirect every total-block path constant to tmp_path for one test."""
built = build_paths(tmp_path)
with (
patch(f"{BLOCK}.TOTAL_BLOCK_LOCK_FILE", built.lock_file),
patch(f"{IPTABLES}._IPTABLES_IP_CACHE_FILE", built.ip_cache_file),
patch(f"{HOSTS}.HOSTS_FILE", built.hosts_file),
patch(f"{PURGE}._STEAM_PURGE_LOG_FILE", built.purge_log_file),
patch(f"{PURGE}._STEAM_REMNANT_PATHS", built.remnant_paths),
):
yield built
# ──────────────────────────────────────────────────────────────
# Live process-table isolation
#
# The Steam-restart guard asks whether a game is running by scanning
# /proc for SteamAppId. Unpatched, that reads the *developer's* real
# process table: running the suite while playing a game made three
# restart tests fail on 2026-08-28. Tests that care about the guard
# patch this themselves.
# ──────────────────────────────────────────────────────────────
@pytest.fixture(autouse=True)
def _no_live_games() -> Iterator[None]:
"""Report no running games unless a test says otherwise."""
with patch(
"steam_backlog_enforcer._steam_restart_guard.get_running_steam_game_pids",
return_value={},
):
yield