Files
iris_x_hermes/gateway-plugin/outbox.py
T
ARIA 29d0c1a73f
CI / Gateway plugin tests (push) Successful in 5m46s
CI / Kotlin tests (android host + desktop) (push) Successful in 6m57s
Fix two gateway test failures: outbox lane scoping + SSE teardown race
- outbox: delete_message/message_info now match the exact lane first
  (a flat-lane delete/lookup with thread_id=None sees only frames with
  no thread_id) and fall back to the message_id across all lanes only
  when the exact lane matches nothing. Previously lane=None meant
  'any lane' in the first pass, so a flat-lane delete also removed
  same-id frames from threads (test expected 3 removed, got 4).

- http_server: the SSE live loop skipped queued frames when stop() set
  sub.closed before the handler thread reached the loop (descheduled
  under load between the initial hello/status writes and the loop).
  The loop now drains frames queued before the close, so the
  status{restarting} teardown broadcast always reaches the client
  before EOF (test_disconnect_broadcasts_status_restarting was flaky
  ~70% under CPU load).
2026-08-23 14:45:09 +02:00

467 lines
20 KiB
Python

"""SQLite offline outbox + monotonic sync cursor (M3).
Undelivered frames (sent while no device is live) are appended with a
**monotonic** cursor so a reconnecting app can ``sync {cursor}`` the delta
without re-reading full history. The cursor is a separate high-water counter
that only ever increases -- pruning old rows never resets it, so a late
reconnect can't be handed a cursor lower than one it already saw.
Retention prunes rows older than ``outbox_retention_hours`` (default 72h). A
device offline longer than the window misses those frames; it recovers full
context via ``history`` (M5 wires push so the device is woken to sync).
Storage: ``get_hermes_home()/"iris"/outbox.db``.
Milestone M3 (built), extended in M5 (push integration).
"""
import contextlib
import json
import logging
import sqlite3
import threading
import time
from pathlib import Path
from typing import Any
logger = logging.getLogger(__name__)
DEFAULT_RETENTION_HOURS = 72
_REPLAY_LIMIT = 1000
_PRUNE_INTERVAL_S = 3600.0
# Row cap (docs/08 §8.3): a very busy offline period must not grow the outbox
# without bound; the oldest rows beyond the cap are pruned and the app is
# told (generic notification, throttled in the adapter).
DEFAULT_MAX_ROWS = 5000
def _frame_thread_id(frame: dict[str, Any]) -> str | None:
"""A frame's lane: its ``thread_id``, normalized (absent/blank -> None)."""
tid = frame.get("thread_id")
if not isinstance(tid, str) or not tid.strip():
return None
return tid
class Outbox:
"""Persistent outbox under ``get_hermes_home()/"iris"``.
Thread-safe (single connection + lock); operations are small and fast
enough to run inline on the gateway's asyncio loop (mirrors
``DeviceRegistry`` / ``ChannelDirectory``).
"""
def __init__(
self,
db_path: Path,
retention_hours: int = DEFAULT_RETENTION_HOURS,
max_rows: int = DEFAULT_MAX_ROWS,
):
self._db_path = Path(db_path)
self._db_path.parent.mkdir(parents=True, exist_ok=True)
self._retention_hours = max(1, int(retention_hours))
self._max_rows = max(1, int(max_rows))
self._lock = threading.Lock()
self._last_prune = 0.0
self._overflow_pruned = 0
self._conn = sqlite3.connect(str(self._db_path), check_same_thread=False)
self._conn.row_factory = sqlite3.Row
with self._lock:
self._conn.execute("PRAGMA journal_mode=WAL")
self._conn.execute(
"""
CREATE TABLE IF NOT EXISTS outbox (
cursor INTEGER PRIMARY KEY,
chat_id TEXT,
frame TEXT NOT NULL,
created REAL NOT NULL DEFAULT 0
)
"""
)
self._conn.execute("CREATE INDEX IF NOT EXISTS idx_outbox_created ON outbox (created)")
self._conn.execute(
"""
CREATE TABLE IF NOT EXISTS counters (
name TEXT PRIMARY KEY,
value INTEGER NOT NULL DEFAULT 0
)
"""
)
self._conn.commit()
# ── append / cursor ───────────────────────────────────────────────────
def append(self, chat_id: str | None, frame_json: str) -> int:
"""Append a frame; returns the (monotonic) cursor assigned to it."""
now = time.time()
with self._lock:
self._conn.execute(
"INSERT INTO counters (name, value) VALUES ('cursor', 1) "
"ON CONFLICT(name) DO UPDATE SET value = value + 1"
)
row = self._conn.execute("SELECT value FROM counters WHERE name = 'cursor'").fetchone()
cursor = int(row["value"]) if row else 1
self._conn.execute(
"INSERT INTO outbox (cursor, chat_id, frame, created) VALUES (?, ?, ?, ?)",
(cursor, chat_id, frame_json, now),
)
self._enforce_row_cap()
self._conn.commit()
self._maybe_prune()
return cursor
def take_overflow_pruned(self) -> int:
"""Rows pruned by the row cap since the last call (and reset to 0).
The adapter turns a non-zero count into a (throttled) generic
notification so the app knows older frames are gone.
"""
with self._lock:
n = self._overflow_pruned
self._overflow_pruned = 0
return n
def _enforce_row_cap(self) -> None:
"""Drop the oldest rows beyond ``max_rows`` (caller holds the lock)."""
row = self._conn.execute("SELECT COUNT(*) AS n FROM outbox").fetchone()
n = int(row["n"]) if row else 0
excess = n - self._max_rows
if excess <= 0:
return
self._conn.execute(
"DELETE FROM outbox WHERE cursor IN ("
" SELECT cursor FROM outbox ORDER BY cursor ASC LIMIT ?)",
(excess,),
)
self._overflow_pruned += excess
logger.info("iris outbox: row cap pruned %s oldest row(s)", excess)
def latest_cursor(self) -> int:
"""The high-water cursor (0 when nothing has been appended)."""
with self._lock:
row = self._conn.execute("SELECT value FROM counters WHERE name = 'cursor'").fetchone()
return int(row["value"]) if row else 0
# ── replay ────────────────────────────────────────────────────────────
def replay(self, cursor: int, limit: int = _REPLAY_LIMIT) -> list[dict[str, Any]]:
"""Frames with ``cursor > `cursor```, oldest first.
Each entry: ``{cursor, chat_id, frame}`` where ``frame`` is the parsed
frame dict (the caller re-serializes / forwards it to the device).
"""
cursor = max(0, int(cursor or 0))
limit = max(1, min(int(limit or _REPLAY_LIMIT), _REPLAY_LIMIT))
with self._lock:
rows = self._conn.execute(
"SELECT cursor, chat_id, frame FROM outbox "
"WHERE cursor > ? ORDER BY cursor ASC LIMIT ?",
(cursor, limit),
).fetchall()
out: list[dict[str, Any]] = []
for r in rows:
try:
frame = json.loads(r["frame"])
except (json.JSONDecodeError, TypeError):
continue
if not isinstance(frame, dict):
continue
out.append({"cursor": int(r["cursor"]), "chat_id": r["chat_id"], "frame": frame})
return out
# ── history (full message history for a chat/thread) ──────────────────
def history(
self,
chat_id: str,
thread_id: str | None = None,
before_message_id: str | None = None,
limit: int = 50,
) -> dict[str, Any]:
"""Final messages for a chat/thread, for the ``history`` frame.
Reconstructs the message list from the outbox log: a final message is
either a standalone ``message`` frame (user echo / non-streaming
assistant) or a ``message.stop`` frame (streaming assistant final).
Intermediate frames (``message.start``/``message.update``, tool,
commentary, notification, …) are skipped. Messages are deduplicated by
``message_id`` and returned oldest → newest.
Pagination: ``before_message_id`` returns the page of messages older
than that id (the newest page when omitted). Returns
``{messages, has_more, oldest_message_id}``.
"""
limit = max(1, min(int(limit or 50), 200))
with self._lock:
rows = self._conn.execute(
"SELECT cursor, frame FROM outbox WHERE chat_id = ? ORDER BY cursor ASC",
(chat_id,),
).fetchall()
final: list[dict[str, Any]] = []
for r in rows:
try:
frame = json.loads(r["frame"])
except (json.JSONDecodeError, TypeError):
continue
if not isinstance(frame, dict):
continue
# Exact lane match: a flat-lane history (thread_id=None) must NOT
# include frames that belong to a thread, and vice versa. (The old
# loose filter let auto-threaded messages leak into the flat lane
# on restart, where a delete sent with thread_id=None then matched
# nothing in the outbox and the messages "resurrected" later.)
if _frame_thread_id(frame) != thread_id:
continue
ftype = frame.get("type")
payload = frame.get("payload")
if not isinstance(payload, dict):
continue
if ftype == "message":
role = payload.get("role")
if role not in ("user", "assistant"):
continue
msg = {
"cursor": int(r["cursor"]),
"message_id": payload.get("message_id"),
"role": role,
"text": payload.get("text", ""),
"reasoning": payload.get("reasoning"),
"model": payload.get("model"),
"tokens": payload.get("tokens"),
"runtime": payload.get("runtime"),
"ts": payload.get("ts"),
}
# Omit ``media`` when absent (schema: array, not null) — a
# ``"media": null`` would break the app's deserialization.
if payload.get("media"):
msg["media"] = payload["media"]
final.append(msg)
elif ftype == "message.stop":
# Streaming finals carry no media (offers are separate
# frames); omit the key (schema: array, not null).
final.append(
{
"cursor": int(r["cursor"]),
"message_id": payload.get("message_id"),
"role": "assistant",
"text": payload.get("final_text", ""),
"reasoning": payload.get("reasoning"),
"model": payload.get("model"),
"tokens": payload.get("tokens"),
"runtime": payload.get("runtime"),
"ts": payload.get("ts"),
}
)
# Deduplicate by message_id (keep the latest occurrence), keep order.
by_id: dict[str, dict[str, Any]] = {}
for m in final:
mid = m.get("message_id")
if mid:
by_id[mid] = m
ordered = sorted(by_id.values(), key=lambda m: m["cursor"])
# Paginate: messages older than ``before_message_id`` (newest page when
# omitted / not found — a pruned anchor falls back to the newest page).
if before_message_id:
idx = next(
(i for i, m in enumerate(ordered) if m["message_id"] == before_message_id),
None,
)
pool = ordered[:idx] if idx is not None else ordered
else:
pool = ordered
has_more = len(pool) > limit
page = pool[-limit:] if has_more else list(pool)
oldest_message_id = page[0]["message_id"] if page else None
for m in page:
m.pop("cursor", None)
# Omit absent optional fields (the app's serializer treats a
# missing key as its default, but a JSON ``null`` for a
# non-nullable field like ``media`` would fail to parse).
for key in ("reasoning", "model", "tokens", "runtime", "ts", "media"):
if m.get(key) is None:
m.pop(key, None)
return {
"messages": page,
"has_more": has_more,
"oldest_message_id": oldest_message_id,
}
# ── message deletion ──────────────────────────────────────────────────
def message_info(
self,
chat_id: str,
message_id: str,
thread_id: str | None = None,
) -> dict[str, Any] | None:
"""Look up a message's final frame data (role / text / ts) in the outbox.
Used to match a ``message.delete`` to the hermes session-store row
(which is keyed by content + timestamp, not the plugin's message id).
Returns ``{role, text, ts}`` for the message's final frame -- a
standalone ``message`` frame when present, else the ``message.stop``
frame of a streamed reply -- or ``None`` when the message is not in the
outbox (e.g. already pruned by retention).
The lane is matched exactly first (a flat-lane lookup, ``thread_id
= None``, sees only frames with no ``thread_id``); when that finds
nothing the lookup falls back to the ``message_id`` alone across all
lanes (it is a unique uuid4), so a stale/missing ``thread_id`` on the
request still resolves the row.
"""
if not message_id:
return None
with self._lock:
rows = self._conn.execute(
"SELECT frame FROM outbox WHERE chat_id = ?", (chat_id,)
).fetchall()
def scan(lane: str | None, exact: bool) -> dict[str, Any] | None:
msg_frame: dict[str, Any] | None = None
stop_frame: dict[str, Any] | None = None
for r in rows:
try:
frame = json.loads(r["frame"])
except (json.JSONDecodeError, TypeError):
continue
if not isinstance(frame, dict):
continue
if exact and _frame_thread_id(frame) != lane:
continue
payload = frame.get("payload")
if not isinstance(payload, dict) or payload.get("message_id") != message_id:
continue
ftype = frame.get("type")
if ftype == "message":
msg_frame = {
"role": payload.get("role"),
"text": payload.get("text", ""),
"ts": payload.get("ts"),
}
elif ftype == "message.stop":
stop_frame = {
"role": "assistant",
"text": payload.get("final_text", ""),
"ts": payload.get("ts"),
}
return msg_frame or stop_frame
return scan(thread_id, exact=True) or scan(None, exact=False)
def delete_message(
self,
chat_id: str,
message_id: str,
thread_id: str | None = None,
) -> int:
"""Remove every outbox frame belonging to *message_id* in *chat_id*.
A message can span several frames (``message`` / ``message.start`` /
``message.update`` / ``message.stop`` / ``media.offer`` /
``commentary``); all of them are removed so neither ``history`` nor a
``sync`` replay can resurrect the message. The delete is scoped to the
exact lane first: a flat-lane delete (``thread_id=None``) matches only
frames with no ``thread_id``, and a thread delete matches only that
thread's frames. When the exact lane matches nothing, the delete falls
back to the ``message_id`` alone (it is a unique uuid4, so it cannot
hit the wrong message) — this keeps deletes working when the request's
lane is stale or missing (e.g. a message the app cached in the flat
lane that the gateway auto-threaded). Returns the number of rows
removed (0 when the message is not in the outbox — e.g. already pruned
by retention).
"""
if not message_id:
return 0
with self._lock:
rows = self._conn.execute(
"SELECT cursor, frame FROM outbox WHERE chat_id = ?", (chat_id,)
).fetchall()
def cursors_for(lane: str | None, exact: bool) -> list[int]:
out: list[int] = []
for r in rows:
try:
frame = json.loads(r["frame"])
except (json.JSONDecodeError, TypeError):
continue
if not isinstance(frame, dict):
continue
if exact and _frame_thread_id(frame) != lane:
continue
payload = frame.get("payload")
if isinstance(payload, dict) and payload.get("message_id") == message_id:
out.append(int(r["cursor"]))
return out
# Exact lane first (a flat-lane delete must not reach into
# threads); fall back to the message_id across all lanes only
# when the exact lane matches nothing (stale/missing thread_id).
cursors = cursors_for(thread_id, exact=True) or cursors_for(thread_id, exact=False)
if not cursors:
return 0
# One bound-parameter delete per cursor (a message spans only a few
# frames); same transaction, no string-built SQL.
for cursor in cursors:
self._conn.execute("DELETE FROM outbox WHERE cursor = ?", (cursor,))
self._conn.commit()
return len(cursors)
def delete_lane(self, chat_id: str, thread_id: str | None = None) -> int:
"""Remove every outbox frame for a lane (channel or thread).
* ``thread_id is None`` -> a **channel**: all frames whose ``chat_id``
column is *chat_id* (the flat lane plus every thread under it).
* ``thread_id`` set -> a **thread**: frames for *chat_id* whose frame
carries that ``thread_id``.
Called on channel/thread deletion so neither ``history`` nor a ``sync``
replay can resurrect the lane's messages. Returns the number of rows
removed.
"""
with self._lock:
rows = self._conn.execute(
"SELECT cursor, frame FROM outbox WHERE chat_id = ?", (chat_id,)
).fetchall()
cursors: list[int] = []
for r in rows:
if thread_id is None:
cursors.append(int(r["cursor"]))
continue
try:
frame = json.loads(r["frame"])
except (json.JSONDecodeError, TypeError):
continue
if isinstance(frame, dict) and frame.get("thread_id") == thread_id:
cursors.append(int(r["cursor"]))
if not cursors:
return 0
for cursor in cursors:
self._conn.execute("DELETE FROM outbox WHERE cursor = ?", (cursor,))
self._conn.commit()
return len(cursors)
# ── retention ─────────────────────────────────────────────────────────
def _maybe_prune(self) -> None:
now = time.time()
if now - self._last_prune < _PRUNE_INTERVAL_S:
return
self._last_prune = now
cutoff = now - self._retention_hours * 3600
with self._lock:
try:
self._conn.execute("DELETE FROM outbox WHERE created < ?", (cutoff,))
self._conn.commit()
except sqlite3.Error as e:
logger.debug("iris outbox: prune failed: %s", e)
def prune(self) -> None:
"""Force a retention prune (ignores the interval throttle)."""
self._last_prune = 0.0
self._maybe_prune()
def close(self) -> None:
with self._lock, contextlib.suppress(Exception):
# Best-effort: a close failure on shutdown is not actionable.
self._conn.close()