The guard in startup_events() read:
if getenv_compat("FEEDBACK_DEMO_MODE") or getenv_compat("FEEDBACK_DEMO_MODE") == "1" \
and not _DEMO_JANITOR_STARTED:
`and` binds tighter than `or`, so that is `A or (B and C)`. The not-already-started half
never ran when the env var was truthy — the only case that reaches it at all. A second
startup started a SECOND janitor thread, overwrote the handle, and shutdown then joined
only the last: the first leaked and kept firing registered hooks hourly, forever.
The guard now lives INSIDE start_janitor(). A caller cannot get operator precedence wrong
if there is nothing left for it to get wrong.
━━━ THREE WAYS TO WRITE THIS GUARD WRONG. I HIT ALL THREE. ━━━
1. NO GUARD — the original bug. Double-start, orphaned thread.
2. GUARD ON THE FLAG (`if _DEMO_JANITOR_STARTED: return`). Codex [P2]. stop_janitor()
DELIBERATELY leaves that flag True when a hook outruns its join timeout, so that a later
startup cannot spawn a janitor beside a live one. But the hook usually finishes a moment
later: the thread exits and the flag is stale. A flag-keyed guard then refuses to start a
replacement for the rest of the process — demo cleanup silently dead. (The original bug
accidentally MASKED this by always starting.)
3. GUARD ON LIVENESS ALONE (`if thread.is_alive(): return`). Codex [P2], second pass. A
timed-out stop leaves the old thread ALIVE BUT DOOMED — its stop event is set and it
exits as soon as its current hook returns. Treating that as a running janitor skips the
replacement, and we are back at (2) a second later.
So: a janitor counts as running only if its thread is alive AND it has not been told to stop.
━━━ AND EACH JANITOR NOW OWNS ITS STOP EVENT ━━━
start_janitor() used to `_DEMO_JANITOR_STOP.clear()` a single SHARED Event. Start a
replacement while a doomed thread is still finishing a hook and that clear RESURRECTS it: it
loops back to wait(), sees the flag cleared, and carries on. Two janitors — the exact bug we
started from. A fresh Event per janitor makes it impossible; the old thread waits on its own
event, which stays set, so it can only exit.
Env semantics UNCHANGED, verified across every value ("", "1", "0", "true", "false", "off"):
the old expression and demo_mode_enabled() agree on all of them. The only behavioural change
is the idempotency fix.
FOUR tests, and each of the three wrong guards fails a different subset:
no guard -> 2 fail (double start; orphaned thread)
guard on the flag -> 2 fail (never restarts after a timed-out stop)
liveness alone -> 1 fail (no replacement for a doomed janitor)
liveness + not-stopping -> all pass
pytest 2416, pyflakes 0, Codex 0.
Closes#902
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
lib/demo_mode.py (342). server.py 1,870 -> 1,649.
The read-only request guard (its 96-entry blocked-route table + the middleware) and the
hourly session janitor (registry, hook runner, thread). Bodies VERBATIM.
THE MIDDLEWARE NEEDS `app`, SO THE MODULE TAKES IT. _demo_mode_guard is an
@app.middleware("http") and cannot exist without an app object. Rather than have a module
under lib/ reach for a global, it exposes install(app) and server.py — which owns the app —
hands it over. The janitor is symmetrical: start_janitor() / stop_janitor(), called from
server.py's startup and shutdown hooks, where the process lifecycle actually lives.
register_demo_janitor_hook IS PART OF THE PLUGIN CONTRACT. It is a key in plugin_context,
so plugins hold it as a LIVE REFERENCE from setup(). server.py imports this exact object
and puts it in the dict unchanged — identity preserved, and
tests/test_plugin_context_contract.py (#898, merged) fails if that ever stops being true.
This is the first carve that guard has actually protected.
━━━ stop_janitor()'s ORDER IS LOAD-BEARING ━━━
The obvious way to write it — clear the "started" flag, then join — is WRONG, and I wrote
it that way first. server.py's original deliberately returns EARLY, leaving
_DEMO_JANITOR_STARTED True and the thread handle intact, when the thread outlives the join:
# Leave _DEMO_JANITOR_STARTED True so a new janitor is not
# spawned by a subsequent startup while the old one is alive.
Clearing the flag first quietly reintroduces exactly the double-janitor leak the flag
exists to prevent. Preserved byte-for-byte, and the reason is now written down at the
function rather than only at its single call site.
━━━ A BUG MOVED VERBATIM, ON PURPOSE ━━━
if getenv_compat("FEEDBACK_DEMO_MODE") or getenv_compat("FEEDBACK_DEMO_MODE") == "1" \
and not _DEMO_JANITOR_STARTED:
`and` binds tighter than `or`, so this is `A or (B and C)` — the not-already-started
re-entry guard is DEAD whenever the env var is truthy, which is the only case that runs.
A second startup leaks a janitor thread (the handle is overwritten, so shutdown joins only
the last). Verified. Preserved exactly and filed as issue #902: a carve whose whole value
is being provably behaviour-neutral is not the place to change behaviour.
pyflakes caught three more missing imports on the way in (uuid, warnings x2). Five carves,
ten missing imports, every one a NameError on a live path.
pytest 2399, pyflakes 0, Codex 0.
Refs #48
Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
