Warn in the navbar when a hard drive runs hot (#470)

* Warn in the navbar when a hard drive runs hot.

The navbar already warns about high CPU temperature; add the equivalent
for drive temperature, since an overheating disk risks data loss and a
user running an off-grid library won't be watching for it.

- controller/lib/smart.py: build_smart_stats() returns the per-drive list
  plus warning thresholds (55C warn / 65C critical) so the frontend stays
  dumb. Reading SMART wakes a spun-down drive, so drives in standby are
  detected with `smartctl -n standby` (which works through the USB-SATA
  bridges where `hdparm -C` only reports "unknown") and NOT read; their
  last-known reading is reused instead.
- controller/lib/status_worker.py: collect SMART on a slow 60s cadence
  (cached on app.state) rather than the 5s status loop, so the poll does
  not keep USB drives awake. The result flows into the same cached payload
  via a new collect_all_status(smart_stats=...) kwarg.
- Frontend: useDriveTemperature() reports the hottest drive; Nav.js shows
  a yellow/red `hdd` icon with a popup, mirroring the CPU warning.

Verified on hardware: smart_stats reports real temps and SMART runs once
a minute, not every 5s.

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

* Address Greptile review on HDD temperature warning.

- _drive_is_asleep also treats SLEEP mode as asleep, not just STANDBY.
  smartctl -n standby exits "Device is in SLEEP mode, exit(2)" for the
  deeper power mode; matching only "STANDBY" meant a SLEEP drive was read
  and woken — the exact failure the check prevents.  (TDD: added a failing
  SLEEP test, then fixed.)
- useDriveTemperature uses ?? instead of || for the temperature fallback
  to express "fall back only on null/undefined".

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>

controller/conftest.py

@@ -29,6 +29,7 @@ _FAKE_CACHED_STATUS = {
     "processes_stats": [],
     "disk_bandwidth_stats": {},
     "uptime_stats": {"uptime_seconds": 0},
+    "smart_stats": {"drives": [], "high_temperature": 55, "critical_temperature": 65},
     "last_status": "1970-01-01T00:00:00",
 }