signal-bot/signal_bot/bot.py
Krzysztof kuhy Rudnicki 423db74788 feat: survive an outage, and keep the counts that predate the event log
Two gaps kuhy pointed at, both about not losing messages the bot was not
awake for.

Missed *counts* already worked and now have a test saying so: an event
is keyed by Signal's own send timestamp, signal-cli keeps receiving in
json-rpc mode while the bot is down, and the replay on reconnect is
absorbed by the primary key. Messages sent during an outage come back
filed under the day they were actually sent.

Missed *reports* did not. A bot that was down at 21:37 never posted that
day and never went back for it. The last day reported is now written
down per group, so returning from an outage posts every day it owes,
oldest first, bounded at a week. A day nobody spoke on is marked settled
rather than reconsidered forever; a day whose send *failed* is
deliberately left unmarked so the next run retries it.

The pre-0.1.0 counter is carried over rather than discarded: 342
messages across four groups, 319 of them in the real group. What the
counter cannot say is when, so before assuming, four places were checked
for a real inbound timestamp -- journald (commands only), signal-cli's
account.db (19 tables, none inbound; message_send_log is the bot's own
outgoing resend buffer), Signal Desktop (unlinked, empty, and it does
not backfill on link), and signal-api/attachments (the one dated
artifact). The attachment mtimes put 12 files on 29 Aug against 79 on
30 Aug inside the counter's window, which is why the whole counter is
filed under 30 Aug rather than split.

That day is marked in group_meta, and a marked day claims no average
message length -- those messages have a count but no lengths, so
"średnio 0 znaków" would be a measurement nobody took. The mark is also
what will keep the day out of personal records: a merged 22-hour day
would otherwise hand somebody a record they never set.

render_day's options move into a DayView value object; the keyword list
had reached eight and this renderer is about to be shared by several
commands that each want a different subset.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015d48t4qsQLBJnDwDeo2Nom
2026-08-30 19:23:53 +02:00

185 lines
6.6 KiB
Python

# Copyright (c) 2026 Krzysztof Rudnicki. MIT licence.
"""The bot object itself: the command wiring and the two loops it owns.
Split out of app.py, which had grown to the 250-line cap with the ASGI
entrypoint, the object graph and the startup sequence all in one file. This
half is "what the bot does"; startup.py is "how it comes up".
"""
from __future__ import annotations
import datetime as dt
import logging
from collections.abc import Callable
from dataclasses import dataclass
from typing import TYPE_CHECKING
from signal_bot.animals import register_animal_commands
from signal_bot.announce import announce_all
from signal_bot.changelog import load as load_changelog
from signal_bot.decorators import cooldown
from signal_bot.dispatcher import Dispatcher
from signal_bot.http_client import ImageFetcher
from signal_bot.popes_command import register_pope_command
from signal_bot.portraits import PortraitCache
from signal_bot.registry import CommandRegistry
from signal_bot.stats import periods
from signal_bot.stats.backfill import backfilled_day
from signal_bot.stats.nightly import days_to_report, mark_reported
from signal_bot.stats.queries_day import day_rows, day_totals, with_silent
from signal_bot.stats.render_day import DayView, render_day
from signal_bot.versioning import POPE, __version__
if TYPE_CHECKING:
import httpx
from signal_bot.config import Config
from signal_bot.discovery import GroupDirectory
from signal_bot.liveness import Liveness
from signal_bot.store import Store
from signal_bot.transport_base import Transport
logger = logging.getLogger(__name__)
Clock = Callable[[], dt.datetime]
COOLDOWN_MESSAGE = "BEEP BOOP POCZEKAJ {seconds} SEKUND."
def now_local() -> dt.datetime:
"""Current Warsaw time -- the group's day, not the host's.
One definition of "day" for the whole bot. The daily scheduler fires
against this, ``CommandContext.day`` keys the daily limits off it, and
every event is filed under it, so a host in another timezone (or a
container on UTC) cannot make those three disagree.
"""
return periods.now()
@dataclass
class Bot:
"""The wired-up bot and its two long-running loops."""
config: Config
store: Store
transport: Transport
directory: GroupDirectory
dispatcher: Dispatcher
liveness: Liveness
# The same clock the dispatcher files events against. One clock, so
# "which day is it" cannot differ between writing an event and reporting
# on it -- which is precisely the drift the old reset-after-post had.
clock: Clock = now_local
async def publish_report(self) -> None:
"""Post each group's report for the last full day.
Yesterday, not "since the last post". The old scheme reset a counter
after publishing, so a group nobody spoke in was never reset and its
next report silently covered two days. A day is now a property of
when a message was sent, which cannot drift.
Nothing is deleted afterwards. Each group is isolated on purpose: one
group refusing the report -- the bot removed from it, a transient send
failure -- must not cost every other group its report.
"""
latest = periods.previous_day(self.clock())
for group_id in self.directory.groups:
for day in days_to_report(self.store, group_id, latest):
if not await self._report_one(group_id, day):
# Left unmarked on purpose, so the next run retries it
# rather than skipping the day forever.
break
mark_reported(self.store, group_id, day)
async def _report_one(self, group_id: str, day: str) -> bool:
"""Post one group's report for one day. False means try again later."""
rows = day_rows(self.store.events, group_id, day)
if not rows:
# Settled, not pending. The bot is in every group it was ever
# added to, so posting "nobody spoke" would mean a nightly ping
# into every quiet group, forever.
logger.info("no messages in %s on %s; not posting a report", group_id, day)
return True
named = self.store.events.roster(group_id, self.directory.member_identities(group_id))
with_zeros = with_silent(rows, [(p.uuid, p.display_name) for p in named])
text = render_day(
day,
with_zeros,
day_totals(with_zeros),
DayView(
since=self.store.events.first_day(group_id),
approximate=backfilled_day(self.store, group_id) == day,
),
)
try:
await self.transport.send_text(text, group_id)
except Exception:
logger.exception("daily report failed for group %s", group_id)
return False
return True
async def announce_release(self) -> None:
"""Tell every watched group about a version it has not seen yet."""
await announce_all(
self.transport,
self.store,
list(self.directory.groups),
load_changelog(),
__version__,
POPE.name,
)
async def listen(self) -> None:
"""Consume frames forever, handing each to the dispatcher."""
async for frame in self.transport.receive():
# Every frame counts, including receipts and typing indicators:
# the question this answers is "can anything still reach us?",
# not "did anybody address the bot?".
self.liveness.note_envelope()
await self.dispatcher.handle(frame)
def build_bot(
config: Config,
client: httpx.AsyncClient,
store: Store,
transport: Transport,
directory: GroupDirectory,
liveness: Liveness,
clock: Clock = now_local,
) -> Bot:
"""Assemble the object graph from its injected dependencies."""
registry = CommandRegistry()
fetcher = ImageFetcher(client)
rate_limit = cooldown(config.cooldown_seconds, COOLDOWN_MESSAGE)
register_animal_commands(
registry,
fetcher,
decorate=rate_limit,
endpoints=config.image_api_urls,
)
register_pope_command(
registry,
PortraitCache(fetcher, config.commons_api_url),
decorate=rate_limit,
)
dispatcher = Dispatcher(
config=config,
directory=directory,
registry=registry,
store=store,
transport=transport,
clock=clock,
)
return Bot(
config=config,
store=store,
transport=transport,
directory=directory,
dispatcher=dispatcher,
liveness=liveness,
clock=clock,
)