Compare commits

...

9 Commits

Author SHA1 Message Date
lebaudantoine 03dfea3aac 💄(frontend) position the login hint dynamically next to the button
Compute the position of the login hint at render time so it is
always displayed close to the login button, regardless of the
button's placement or the current viewport size.
2026-09-06 23:15:07 +02:00
lebaudantoine 5c8a1e6251 📈(frontend) track missing lobby participant on accept/reject
When a moderator accepts or rejects a lobby entry that no longer
exists, emit a tracking event so we can measure how often it
happens.

This signal will help tune the lobby polling interval: too many
"not found" events means the moderator side is working from a stale
list. Keep raising the error to the client on top of tracking it,
so the frontend still surfaces the issue (its current handling of
this case is still incomplete).
2026-09-06 23:15:07 +02:00
lebaudantoine b687af2bf9 🔊(backend) log request duration in Gunicorn workers
Include the time taken by each request in the Gunicorn worker
access logs, so we can spot slow endpoints and correlate latency
patterns directly from the logs.
2026-09-06 23:15:07 +02:00
lebaudantoine 43963c3d24 ️(backend) refactor presence cache to bound key lookups per room
The previous presence cache lookup keyed off a scan over the whole
cache, so its cost was O(db_size) rather than O(room_size).
Combined with the recent switch to cursor-based `SCAN` at an
inappropriate page size, this caused a lot of Redis round-trips and
noticeably slowed down the backend pods under load.

Refactor the presence cache to keep a per-room set of all its
participant keys. Lookups now iterate that set instead of scanning
the whole database.

Complexity is now bounded by room size, not database size, which
should restore the backend performance to its previous levels while
keeping the lobby behavior unchanged.
2026-09-06 23:15:07 +02:00
lebaudantoine 6ef48f795e ️(backend) refactor lobby storage to bound key lookups per room
The previous lobby lookup keyed off a scan over the whole cache, so
its cost was O(db_size) rather than O(room_size). Combined with the
recent switch to cursor-based `SCAN` at an inappropriate page size,
this caused a lot of Redis round-trips and noticeably slowed down
the backend pods under load.

Refactor the lobby storage to keep a per-room set of all its lobby
keys. Lookups now iterate that set instead of scanning the whole
database:

* Membership in the set acts as a memory of who is supposedly in
  the lobby for a given room.
* Individual keys are then read to check who is actually still
  waiting or accepted.

Complexity is now bounded by room size, not database size, which
should restore the backend performance to its previous levels while
keeping the lobby behavior unchanged.
2026-09-06 23:15:07 +02:00
lebaudantoine ed8abd348b ️(frontend) add trailing slash on the /me endpoint call
The `/me` endpoint was called without a trailing slash, so every
request was going through a 301 redirect before hitting the actual
endpoint.

This endpoint is called by every user at least once per session, so
based on the logs, avoiding the redirect should cut the volume of
requests hitting it by around 10%.
2026-09-06 23:15:06 +02:00
lebaudantoine c2f7399e62 ️(frontend) increase lobby polling interval on both sides
Increase the polling interval used by the lobby feature, on both the
waiting participant side and the moderator side.

The goal is to reduce the volume of requests the lobby generates,
trading a bit of data freshness for better performance.

It will de facto reduce pressure on the backend.

We will observe the impact in production, and revisit these
intervals if the delays turn out to be too aggressive.
2026-09-06 23:15:04 +02:00
lebaudantoine adc74f846c 🐛(frontend) center Avatar initials with a font-aware cap-height ratio
The previous implementation centered initials by measuring `<text>`
with `getBBox()`, which returns the font's advance-width by
ascent-to-descent band, not the ink of the glyphs. On fonts with an
asymmetric band, initials rendered off-center. Marianne is a
particularly clear case: ascent 1131 / descent 256 puts the band
center 87.5/1000 above the caps' optical center, so every avatar
sat ~4.6 viewBox units (~1.5–1.8 px) too low. A follow-up rewrite
using canvas `actualBoundingBox*` metrics fixed it but pulled in a
runtime measurement rig (shared canvas, ref, state, layout effect,
`fonts.load` + `loadingdone`, plus combining-mark stripping) just
to place two uppercase letters.

Since initials are always uppercased, optical centering has a
closed form: `baseline = center + capHeight/2`. Real-world text
fonts have cap heights in a narrow ~0.66–0.73 em band (Marianne is
0.70), so:

    translateY(calc(var(--avatar-cap-height, 0.7) * 0.5em))

is exact for the stock font and within ~0.4 px for any plausible
replacement. No JS, SSR-safe, no first-paint jump, no font-loading
race. Accents float above the cap box instead of dragging the
letter down.

The single font-dependent number remaining (cap height) is exposed
as a CSS variable, so self-hosters overriding the font can override
it next to the font itself, or leave the default. Once
`text-box: trim-both cap alphabetic` ships broadly, even the
variable can go.
2026-09-05 10:43:22 +02:00
lebaudantoine 78ba03a52b 🐛(frontend) restore automatic lower-hand on speaking
Following the re-rendering optimization refactoring, the automatic
lower-hand feature broke: the way `isSpeaking` was read no longer
made sense once we limited how often components in the app
re-render.

Fix the detection so the raised hand is again lowered automatically
when the participant starts speaking, without relying on frequent
re-renders.
2026-09-05 00:29:50 +02:00
20 changed files with 368 additions and 134 deletions
+9
View File
@@ -13,6 +13,8 @@ and this project adheres to
- ✨(frontend) add 1080p sending resolution option #1660
- ✨(backend) add Traefik support via configurable media-auth url header #1649
- ✨(backend) update a room's attributes from the external API
- 🔊(backend) log request duration in Gunicorn workers
- 📈(frontend) track missing lobby participant on accept/reject
### Changed
@@ -22,6 +24,13 @@ and this project adheres to
- 🐛(backend) allow any printable ASCII characters in user sub field #1673
- 🐛(frontend) keep the sending resolution picked while the camera is off #1667
- 🐛(frontend) restore automatic lower-hand on speaking
- 🐛(frontend) center Avatar initials with a font-aware cap-height ratio
- ⚡️(frontend) increase lobby polling interval on both sides
- ⚡️(frontend) add trailing slash on the /me endpoint call
- ⚡️(backend) refactor lobby storage to bound key lookups per room
- ⚡️(backend) refactor presence cache to bound key lookups per room
- 💄(frontend) position the login hint dynamically next to the button
## [1.30.0] - 2026-09-01
+1
View File
@@ -1,5 +1,6 @@
:root {
--fonts-sans: 'Marianne', ui-sans-serif, system-ui, sans-serif;
--avatar-cap-height: 0.7;
}
.Header-beforeLogo {
@@ -14,3 +14,4 @@ accesslog = "-"
# Using '-' for the error log file makes gunicorn log errors to stderr
errorlog = "-"
loglevel = "info"
access_log_format = '%(h)s %(l)s %(u)s %(t)s "%(r)s" %(s)s %(b)s "%(f)s" "%(a)s" %(M)s'
+1
View File
@@ -34,6 +34,7 @@ Let's say you want to change the font of our application to a custom font. You c
:root {
--fonts-sans: 'Roboto', ui-sans-serif, system-ui, sans-serif;
--avatar-cap-height: 0.7;
}
```
+104 -10
View File
@@ -79,6 +79,14 @@ class LobbyService:
Handles participant entry requests, status management, and notifications
using cache for state management and LiveKit for real-time updates.
Participant membership per room is tracked in a native Redis SET (the
"room index") so that listing and clearing a room's lobby never scans
the shared keyspace. The per-participant cache entries remain the source
of truth for state: their TTLs implement liveness (a waiter who stops
polling simply expires). The index only records which participant ids
may exist for a room; a stale id costs one cache miss and is pruned
lazily.
"""
@staticmethod
@@ -86,6 +94,61 @@ class LobbyService:
"""Generate cache key for participant(s) data."""
return f"{settings.LOBBY_KEY_PREFIX}_{room_id!s}_{participant_id}"
@staticmethod
def _get_index_key(room_id: UUID) -> str:
"""Raw Redis key of the per-room participant index (a native SET).
Built through django-redis' make_key so it lives under the same
KEY_PREFIX/version namespace as the participant entries.
"""
return cache.client.make_key(f"{settings.LOBBY_KEY_PREFIX}-index_{room_id!s}")
@staticmethod
def _redis(write: bool = True):
"""Raw redis-py client.
SADD/SREM/SMEMBERS are not exposed by the Django cache API; this is
the documented django-redis escape hatch.
"""
return cache.client.get_client(write=write)
def _index_add(self, room_id: UUID, participant_id: str) -> None:
"""Record a participant id in the room index.
Refreshes a backstop TTL on the index so an abandoned room cannot
leak its set beyond the longest participant timeout.
"""
index_key = self._get_index_key(room_id)
pipe = self._redis().pipeline(transaction=False)
pipe.sadd(index_key, participant_id)
pipe.expire(index_key, settings.LOBBY_ACCEPTED_TIMEOUT)
pipe.execute()
def _index_members(self, room_id: UUID) -> List[str]:
"""All participant ids currently indexed for the room."""
members = self._redis(write=False).smembers(self._get_index_key(room_id))
return [
member.decode() if isinstance(member, bytes) else member
for member in members
]
def _index_touch(self, room_id: UUID) -> None:
"""Re-arm the room index backstop TTL.
Called whenever a participant entry is written or refreshed so the
index always outlives every entry it references — including a lone
waiter whose rolling WAITING TTL would otherwise outlast the
backstop set at enter() time.
"""
self._redis().expire(
self._get_index_key(room_id), settings.LOBBY_ACCEPTED_TIMEOUT
)
def _index_remove(self, room_id: UUID, *participant_ids: str) -> None:
"""Drop participant ids from the room index."""
if participant_ids:
self._redis().srem(self._get_index_key(room_id), *participant_ids)
@staticmethod
def _get_or_create_participant_id(request) -> str:
"""Extract unique participant identifier from the request."""
@@ -209,14 +272,16 @@ class LobbyService:
cache.touch(
self._get_cache_key(room_id, participant_id), settings.LOBBY_WAITING_TIMEOUT
)
self._index_touch(room_id)
def enter(
self, room_id: UUID, participant_id: str, username: str
) -> LobbyParticipant:
"""Add participant to waiting lobby.
Create a new participant entry in waiting status and notify room
participants of the new entry request.
Create a new participant entry in waiting status, index the
participant id for the room, and notify room participants of the
new entry request.
"""
color = utils.generate_color(participant_id)
@@ -245,6 +310,7 @@ class LobbyService:
participant.to_dict(),
timeout=settings.LOBBY_WAITING_TIMEOUT,
)
self._index_add(room_id, participant_id)
return participant
@@ -267,15 +333,31 @@ class LobbyService:
return None
def list_waiting_participants(self, room_id: UUID) -> List[dict]:
"""List all waiting participants for a room."""
"""List all waiting participants for a room.
pattern = self._get_cache_key(room_id, "*")
keys = list(cache.iter_keys(pattern, itersize=utils.CACHE_SCAN_ITERSIZE))
Reads the per-room index (O(participants of this room)) instead of
scanning the shared keyspace. Indexed ids whose cache entry has
expired are pruned lazily here: the entry TTL is the liveness
protocol, so a missing entry means the participant is gone.
"""
if not keys:
member_ids = self._index_members(room_id)
if not member_ids:
return []
data = cache.get_many(keys)
keys_by_id = {
participant_id: self._get_cache_key(room_id, participant_id)
for participant_id in member_ids
}
data = cache.get_many(list(keys_by_id.values()))
dead_ids = [
participant_id
for participant_id, cache_key in keys_by_id.items()
if cache_key not in data
]
self._index_remove(room_id, *dead_ids)
waiting_participants = []
for cache_key, raw_participant in data.items():
@@ -341,16 +423,28 @@ class LobbyService:
participant.status = status
cache.set(cache_key, participant.to_dict(), timeout=timeout)
self._index_touch(room_id)
def clear_room_cache(self, room_id: UUID) -> None:
"""Clear all participant entries from the cache for a specific room."""
"""Clear all participant entries from the cache for a specific room.
cache.delete_pattern(
self._get_cache_key(room_id, "*"), itersize=utils.CACHE_SCAN_ITERSIZE
Deletes the indexed participant entries and the index itself with
targeted commands instead of a full-keyspace pattern scan.
"""
member_ids = self._index_members(room_id)
if member_ids:
cache.delete_many(
[
self._get_cache_key(room_id, participant_id)
for participant_id in member_ids
]
)
self._redis().delete(self._get_index_key(room_id))
def clear_participant_cache(self, room_id: UUID, participant_id: str) -> None:
"""Clear a given participant entry from the cache for a specific room."""
cache_key = self._get_cache_key(room_id, participant_id)
cache.delete(cache_key)
self._index_remove(room_id, participant_id)
+47 -20
View File
@@ -1,25 +1,11 @@
"""Presence cache.
Redis-backed memo of "this identity is currently connected to this room".
This module is intentionally a *pure cache store* with no dependency on other
services, so that `participants_management` (which talks to LiveKit) can
import it without creating an import cycle. The composition of "check cache,
fall back to LiveKit" lives in
`ParticipantsManagement.check_if_in_meeting_cached`.
Only positive answers are stored: a sticky negative would lock out someone
who joins right after a miss for the whole TTL. The TTL is a safety net in
case an invalidation webhook is lost.
"""
"""Presence cache."""
from typing import List
from uuid import UUID
from django.conf import settings
from django.core.cache import cache
from core.utils import CACHE_SCAN_ITERSIZE
class PresenceCache:
"""Store and invalidate (room, identity) presence entries."""
@@ -29,24 +15,65 @@ class PresenceCache:
"""Cache key for a (room, identity) presence entry."""
return f"{settings.PRESENCE_KEY_PREFIX}_{room_id!s}_{identity}"
@staticmethod
def _get_index_key(room_id: UUID | str) -> str:
"""Raw Redis key of the per-room identity index (a native SET).
Built through django-redis' make_key so it lives under the same
KEY_PREFIX/version namespace as the presence entries.
"""
return cache.client.make_key(
f"{settings.PRESENCE_KEY_PREFIX}-index_{room_id!s}"
)
@staticmethod
def _redis(write: bool = True):
"""Raw redis-py client.
SADD/SREM/SMEMBERS are not exposed by the Django cache API; this is
the documented django-redis escape hatch.
"""
return cache.client.get_client(write=write)
def _index_members(self, room_id: UUID | str) -> List[str]:
"""All identities currently indexed for the room."""
members = self._redis(write=False).smembers(self._get_index_key(room_id))
return [
member.decode() if isinstance(member, bytes) else member
for member in members
]
def is_marked_present(self, room_id: UUID | str, identity: str) -> bool:
"""Return True if a positive presence entry exists in cache."""
return bool(cache.get(self._get_cache_key(room_id, identity)))
def mark_present(self, room_id: UUID | str, identity: str) -> None:
"""Record that `identity` is in `room_id`."""
"""Record that `identity` is in `room_id` and index it for the room."""
cache.set(
self._get_cache_key(room_id, identity),
True,
timeout=settings.PRESENCE_CACHE_TIMEOUT,
)
index_key = self._get_index_key(room_id)
pipe = self._redis().pipeline(transaction=False)
pipe.sadd(index_key, identity)
pipe.expire(index_key, settings.PRESENCE_CACHE_TIMEOUT)
pipe.execute()
def clear(self, room_id: UUID | str, identity: str) -> None:
"""Forget presence for one participant (e.g. on participant_left)."""
cache.delete(self._get_cache_key(room_id, identity))
self._redis().srem(self._get_index_key(room_id), identity)
def clear_room(self, room_id: UUID | str) -> None:
"""Forget presence for every participant of a room (on room_finished)."""
cache.delete_pattern(
self._get_cache_key(room_id, "*"), itersize=CACHE_SCAN_ITERSIZE
"""Forget presence for every participant of a room (on room_finished).
Deletes the indexed entries and the index itself with targeted
commands instead of a full-keyspace pattern scan.
"""
identities = self._index_members(room_id)
if identities:
cache.delete_many(
[self._get_cache_key(room_id, identity) for identity in identities]
)
self._redis().delete(self._get_index_key(room_id))
@@ -589,6 +589,9 @@ def test_list_waiting_participants_success(settings):
"color": "#654321",
},
)
lobby_service = LobbyService()
lobby_service._index_add(room.id, "2f7f162f-e7d1-421b-90e7-02bfbfbf8def")
lobby_service._index_add(room.id, "f4ca3ab8a6c04ad88097b8da33f60f10")
response = client.get(f"/api/v1.0/rooms/{room.id}/waiting-participants/")
+126 -17
View File
@@ -2,7 +2,7 @@
Test lobby service.
"""
# pylint: disable=W0621,W0613, W0212, R0913
# pylint: disable=W0621,W0613, W0212, R0913, C0302
# ruff: noqa: PLR0913, PLR0917
import uuid
@@ -24,7 +24,6 @@ from core.services.lobby import (
LobbyParticipantStatus,
LobbyService,
)
from core.services.presence import CACHE_SCAN_ITERSIZE
from core.utils import NotificationError
pytestmark = pytest.mark.django_db
@@ -466,18 +465,22 @@ def test_request_entry_participant_with_role(
def test_refresh_waiting_status(mock_cache, lobby_service, participant_id):
"""Test refreshing waiting status for a participant."""
lobby_service._get_cache_key = mock.Mock(return_value="mocked_cache_key")
lobby_service._index_touch = mock.Mock()
room = RoomFactory(access_level=RoomAccessLevel.RESTRICTED)
lobby_service.refresh_waiting_status(room.id, participant_id)
mock_cache.touch.assert_called_once_with(
"mocked_cache_key", settings.LOBBY_WAITING_TIMEOUT
)
lobby_service._index_touch.assert_called_once_with(room.id)
# pylint: disable=R0917
@mock.patch("core.services.lobby.cache")
@mock.patch("core.utils.generate_color")
@mock.patch("core.utils.notify_participants")
@mock.patch("core.services.lobby.LobbyService._index_add")
def test_enter_success(
mock_index_add,
mock_notify,
mock_generate_color,
mock_cache,
@@ -508,13 +511,16 @@ def test_enter_success(
mock_notify.assert_called_once_with(
room_name=str(room.pk), notification_data={"type": "participantWaiting"}
)
mock_index_add.assert_called_once_with(room.id, participant_id)
# pylint: disable=R0917
@mock.patch("core.services.lobby.cache")
@mock.patch("core.utils.generate_color")
@mock.patch("core.utils.notify_participants")
@mock.patch("core.services.lobby.LobbyService._index_add")
def test_enter_with_notification_error(
mock_index_add,
mock_notify,
mock_generate_color,
mock_cache,
@@ -541,6 +547,7 @@ def test_enter_with_notification_error(
participant.to_dict(),
timeout=settings.LOBBY_WAITING_TIMEOUT,
)
mock_index_add.assert_called_once_with(room.id, participant_id)
@mock.patch("core.services.lobby.cache")
@@ -579,14 +586,15 @@ def test_get_participant_parsing_error(
@mock.patch("core.services.lobby.cache")
def test_list_waiting_participants_empty(mock_cache, lobby_service):
"""Test listing waiting participants when none exist."""
mock_cache.iter_keys.return_value = []
room = RoomFactory(access_level=RoomAccessLevel.RESTRICTED)
lobby_service._index_members = mock.Mock(return_value=[])
lobby_service._index_remove = mock.Mock()
result = lobby_service.list_waiting_participants(room.id)
assert result == []
pattern = f"{settings.LOBBY_KEY_PREFIX}_{room.id!s}_*"
mock_cache.iter_keys.assert_called_once_with(pattern, itersize=CACHE_SCAN_ITERSIZE)
lobby_service._index_members.assert_called_once_with(room.id)
lobby_service._index_remove.assert_not_called()
mock_cache.get_many.assert_not_called()
@@ -595,7 +603,8 @@ def test_list_waiting_participants(mock_cache, lobby_service, participant_dict):
"""Test listing waiting participants with valid data."""
room = RoomFactory(access_level=RoomAccessLevel.RESTRICTED)
cache_key = f"{settings.LOBBY_KEY_PREFIX}_{room.id!s}_participant1"
mock_cache.iter_keys.return_value = [cache_key]
lobby_service._index_members = mock.Mock(return_value=["participant1"])
lobby_service._index_remove = mock.Mock()
mock_cache.get_many.return_value = {cache_key: participant_dict}
result = lobby_service.list_waiting_participants(room.id)
@@ -603,8 +612,8 @@ def test_list_waiting_participants(mock_cache, lobby_service, participant_dict):
assert len(result) == 1
assert result[0]["status"] == "waiting"
assert result[0]["username"] == "test-username"
pattern = f"{settings.LOBBY_KEY_PREFIX}_{room.id!s}_*"
mock_cache.iter_keys.assert_called_once_with(pattern, itersize=CACHE_SCAN_ITERSIZE)
lobby_service._index_members.assert_called_once_with(room.id)
lobby_service._index_remove.assert_called_once_with(room.id)
mock_cache.get_many.assert_called_once_with([cache_key])
@@ -629,7 +638,10 @@ def test_list_waiting_participants_multiple(mock_cache, lobby_service):
"color": "#654321",
}
mock_cache.iter_keys.return_value = [cache_key1, cache_key2]
lobby_service._index_members = mock.Mock(
return_value=["participant1", "participant2"]
)
lobby_service._index_remove = mock.Mock()
mock_cache.get_many.return_value = {
cache_key1: participant1,
cache_key2: participant2,
@@ -646,8 +658,7 @@ def test_list_waiting_participants_multiple(mock_cache, lobby_service):
# Verify all participants have waiting status
assert all(p["status"] == "waiting" for p in result)
pattern = f"{settings.LOBBY_KEY_PREFIX}_{room.id!s}_*"
mock_cache.iter_keys.assert_called_once_with(pattern, itersize=CACHE_SCAN_ITERSIZE)
lobby_service._index_members.assert_called_once_with(room.id)
mock_cache.get_many.assert_called_once_with([cache_key1, cache_key2])
@@ -656,7 +667,8 @@ def test_list_waiting_participants_corrupted_data(mock_cache, lobby_service):
"""Test listing waiting participants with corrupted data."""
room = RoomFactory(access_level=RoomAccessLevel.RESTRICTED)
cache_key = f"{settings.LOBBY_KEY_PREFIX}_{room.id!s}_participant1"
mock_cache.iter_keys.return_value = [cache_key]
lobby_service._index_members = mock.Mock(return_value=["participant1"])
lobby_service._index_remove = mock.Mock()
mock_cache.get_many.return_value = {cache_key: {"invalid": "data"}}
result = lobby_service.list_waiting_participants(room.id)
@@ -681,7 +693,10 @@ def test_list_waiting_participants_partially_corrupted(mock_cache, lobby_service
corrupted_participant = {"invalid": "data"}
mock_cache.iter_keys.return_value = [cache_key1, cache_key2]
lobby_service._index_members = mock.Mock(
return_value=["participant1", "participant2"]
)
lobby_service._index_remove = mock.Mock()
mock_cache.get_many.return_value = {
cache_key1: corrupted_participant,
cache_key2: valid_participant,
@@ -699,8 +714,6 @@ def test_list_waiting_participants_partially_corrupted(mock_cache, lobby_service
mock_cache.delete.assert_called_once_with(cache_key1)
# Verify both cache keys were queried
pattern = f"{settings.LOBBY_KEY_PREFIX}_{room.id!s}_*"
mock_cache.iter_keys.assert_called_once_with(pattern, itersize=CACHE_SCAN_ITERSIZE)
mock_cache.get_many.assert_called_once_with([cache_key1, cache_key2])
@@ -724,7 +737,10 @@ def test_list_waiting_participants_non_waiting(mock_cache, lobby_service):
"color": "#654321",
}
mock_cache.iter_keys.return_value = [cache_key1, cache_key2]
lobby_service._index_members = mock.Mock(
return_value=["participant1", "participant2"]
)
lobby_service._index_remove = mock.Mock()
mock_cache.get_many.return_value = {
cache_key1: participant1,
cache_key2: participant2,
@@ -820,6 +836,7 @@ def test_update_participant_status_success(mock_cache, lobby_service, participan
mock_cache.get.return_value = participant_dict
lobby_service._get_cache_key = mock.Mock(return_value="mocked_cache_key")
lobby_service._index_touch = mock.Mock()
lobby_service._update_participant_status(
room.id,
@@ -837,6 +854,7 @@ def test_update_participant_status_success(mock_cache, lobby_service, participan
mock_cache.set.assert_called_once_with(
"mocked_cache_key", expected_data, timeout=60
)
lobby_service._index_touch.assert_called_once_with(room.id)
lobby_service._get_cache_key.assert_called_once_with(room.id, participant_id)
@@ -881,9 +899,13 @@ def test_clear_room_cache(settings, lobby_service):
timeout=settings.LOBBY_DENIED_TIMEOUT,
)
for participant_id in ("participant1", "participant2", "participant3"):
lobby_service._index_add(room_id, participant_id)
lobby_service.clear_room_cache(room_id)
assert cache.keys(f"test-lobby_{room_id!s}_*") == []
assert lobby_service._index_members(room_id) == []
def test_clear_room_empty(settings, lobby_service):
@@ -910,10 +932,13 @@ def test_clear_participant_cache(lobby_service):
"color": "#123456",
}
cache.set(cache_key, participant_data, timeout=settings.LOBBY_WAITING_TIMEOUT)
lobby_service._index_add(room_id, participant_id)
assert cache.get(cache_key) is not None
assert participant_id in lobby_service._index_members(room_id)
lobby_service.clear_participant_cache(room_id, participant_id)
assert cache.get(cache_key) is None
assert participant_id not in lobby_service._index_members(room_id)
def test_clear_participant_cache_nonexistent(lobby_service):
@@ -927,3 +952,87 @@ def test_clear_participant_cache_nonexistent(lobby_service):
lobby_service.clear_participant_cache(room_id, participant_id)
assert cache.get(cache_key) is None
# Room index integration tests (real cache backend)
def test_index_add_members_remove_roundtrip(lobby_service):
"""The room index records, lists and forgets participant ids."""
room_id = uuid.uuid4()
assert lobby_service._index_members(room_id) == []
lobby_service._index_add(room_id, "participant1")
lobby_service._index_add(room_id, "participant2")
assert sorted(lobby_service._index_members(room_id)) == [
"participant1",
"participant2",
]
# The index carries a backstop TTL so abandoned rooms cannot leak it.
ttl = lobby_service._redis().ttl(lobby_service._get_index_key(room_id))
assert 0 < ttl <= settings.LOBBY_ACCEPTED_TIMEOUT
lobby_service._index_remove(room_id, "participant1")
assert lobby_service._index_members(room_id) == ["participant2"]
@mock.patch("core.utils.notify_participants")
def test_enter_registers_participant_in_room_index(
mock_notify, lobby_service, participant_id, username
):
"""Entering the lobby must index the participant id for the room."""
room_id = uuid.uuid4()
lobby_service.enter(room_id, participant_id, username)
assert lobby_service._index_members(room_id) == [participant_id]
def test_list_waiting_participants_prunes_stale_index_ids(settings, lobby_service):
"""Indexed ids whose cache entry expired are pruned and not listed."""
settings.LOBBY_KEY_PREFIX = "test-lobby-prune"
room_id = uuid.uuid4()
cache.set(
f"test-lobby-prune_{room_id!s}_participant1",
{
"id": "participant1",
"username": "user1",
"status": "waiting",
"color": "#123456",
},
timeout=100,
)
lobby_service._index_add(room_id, "participant1")
# participant2 is indexed but its cache entry has expired.
lobby_service._index_add(room_id, "participant2")
result = lobby_service.list_waiting_participants(room_id)
assert [participant["id"] for participant in result] == ["participant1"]
assert lobby_service._index_members(room_id) == ["participant1"]
def test_refresh_waiting_status_rearms_room_index_ttl(lobby_service, participant_id):
"""A lone waiter's polling must keep the room index alive.
Regression test: the index backstop TTL is only armed at enter() time,
so a participant whose rolling WAITING refreshes outlast it would keep
their entry alive while silently vanishing from the moderator list.
Refreshing the waiting status must therefore re-arm the index TTL.
"""
room_id = uuid.uuid4()
lobby_service._index_add(room_id, participant_id)
index_key = lobby_service._get_index_key(room_id)
redis_client = lobby_service._redis()
redis_client.expire(index_key, 10)
assert redis_client.ttl(index_key) <= 10
lobby_service.refresh_waiting_status(room_id, participant_id)
assert redis_client.ttl(index_key) > 10
assert lobby_service._index_members(room_id) == [participant_id]
@@ -90,19 +90,18 @@ def test_presence_clear_and_clear_room():
assert presence.is_marked_present(other_room, "a") is True
def test_presence_clear_room_scans_in_pages():
"""clear_room removes every match, even across several SCAN pages,
and only within the room."""
def test_presence_clear_room_removes_many_entries_and_the_index():
"""clear_room removes every entry of the room through the index — never
a keyspace scan — and leaves other rooms untouched."""
room_id, other_room = str(uuid4()), str(uuid4())
presence = PresenceCache()
for i in range(7):
presence.mark_present(room_id, f"user-{i}")
presence.mark_present(other_room, "user-0")
# An itersize smaller than the match count forces delete_pattern to
# page through several SCAN cursors rather than finish in one pass.
with mock.patch("core.utils.CACHE_SCAN_ITERSIZE", 3):
presence.clear_room(room_id)
assert all(not presence.is_marked_present(room_id, f"user-{i}") for i in range(7))
assert presence.is_marked_present(other_room, "user-0") is True
assert presence._index_members(room_id) == []
assert presence._index_members(other_room) == ["user-0"]
-3
View File
@@ -499,6 +499,3 @@ def build_telephony_config():
"default_country": country,
"international_phone_number": international,
}
CACHE_SCAN_ITERSIZE = 500
+1 -1
View File
@@ -865,7 +865,7 @@ class Base(Configuration):
"room_lobby", environ_name="LOBBY_KEY_PREFIX", environ_prefix=None
)
LOBBY_WAITING_TIMEOUT = values.PositiveIntegerValue(
3, environ_name="LOBBY_WAITING_TIMEOUT", environ_prefix=None
6, environ_name="LOBBY_WAITING_TIMEOUT", environ_prefix=None
)
LOBBY_DENIED_TIMEOUT = values.PositiveIntegerValue(
5, environ_name="LOBBY_DENIED_TIMEOUT", environ_prefix=None
+6 -42
View File
@@ -1,5 +1,5 @@
import { css, cva, RecipeVariantProps } from '@/styled-system/css'
import React, { useLayoutEffect, useMemo } from 'react'
import React, { useMemo } from 'react'
const avatar = cva({
base: {
@@ -28,24 +28,17 @@ const avatar = cva({
},
})
// Instantiating a segmenter is expensive; create it once and reuse it.
const graphemeSegmenter =
typeof Intl !== 'undefined' && 'Segmenter' in Intl
? new Intl.Segmenter(undefined, { granularity: 'grapheme' })
: undefined
/**
* Returns the first user-perceived character. Some Unicode characters span
* multiple UTF-16 code units, so a naive index into the string can split them
* and yield a broken glyph.
*/
const getFirstGrapheme = (value: string): string => {
if (!value) return ''
if (graphemeSegmenter) {
const [first] = graphemeSegmenter.segment(value)
return first?.segment ?? ''
}
// Fallback: keeps single code points intact (including surrogate pairs).
return Array.from(value)[0] ?? ''
}
@@ -66,36 +59,6 @@ export type AvatarProps = React.HTMLAttributes<HTMLDivElement> & {
export const Avatar = React.memo(
({ name, bgColor, context, notification, style, ...props }: AvatarProps) => {
const initials = useMemo(() => getInitials(name), [name])
const textRef = React.useRef<SVGTextElement>(null)
const [offsetY, setOffsetY] = React.useState(0)
// Optically center the initials: measure the ink bounding box of the
// rendered glyphs and shift them so the box's center sits at the middle
// of the viewBox. Works for any font, weight or glyph shape, unlike a
// hand-tuned dy offset. getBBox() is in local (pre-transform)
// coordinates, so applying the translation never changes the measure.
useLayoutEffect(() => {
const text = textRef.current
if (!text) return
const center = () => {
const box = text.getBBox()
// A hidden element measures as an empty box; keep the default then.
if (box.height === 0) return
setOffsetY(50 - (box.y + box.height / 2))
}
center()
// Glyph metrics can change once webfonts finish loading.
let cancelled = false
document.fonts?.ready.then(() => {
if (!cancelled) center()
})
return () => {
cancelled = true
}
}, [initials])
return (
<div
style={{ backgroundColor: bgColor, ...style }}
@@ -108,15 +71,16 @@ export const Avatar = React.memo(
className={css({ width: '100%', height: '100%', display: 'block' })}
>
<text
ref={textRef}
x="50"
y="50"
transform={`translate(0 ${offsetY})`}
y={50}
textAnchor="middle"
dominantBaseline="central"
fontSize="52"
fontWeight="500"
fill="currentColor"
className={css({
transform:
'translateY(calc(var(--avatar-cap-height, 0.7) * 0.5em))',
})}
>
{initials}
</text>
@@ -18,7 +18,7 @@ export const fetchUser = (
}
): Promise<ApiUser | false> => {
return new Promise((resolve, reject) => {
fetchApi<ApiUser>('/users/me')
fetchApi<ApiUser>('/users/me/')
.then(resolve)
.catch((error) => {
// we assume that a 401 means the user is not logged in
@@ -1,5 +1,6 @@
import { ApiError } from '@/api/ApiError'
import { fetchApi } from '@/api/fetchApi'
import { captureEvent } from '@/features/analytics/telemetry'
import { useMutation, type UseMutationOptions } from '@tanstack/react-query'
export interface EnterRoomParams {
@@ -17,6 +18,7 @@ export const enterRoom = async ({
allowEntry,
participantId,
}: EnterRoomParams): Promise<EnterRoomResponse> => {
try {
return await fetchApi<EnterRoomResponse>(`/rooms/${roomId}/enter/`, {
method: 'POST',
body: JSON.stringify({
@@ -24,6 +26,15 @@ export const enterRoom = async ({
allow_entry: allowEntry,
}),
})
} catch (error) {
if (error instanceof ApiError && error.statusCode === 404) {
captureEvent('lobby_entry_participant_gone', {
room_id: roomId,
allow_entry: allowEntry,
})
}
throw error
}
}
export function useEnterRoom(
@@ -12,8 +12,8 @@ import { keys } from '@/api/queryKeys'
import { queryClient } from '@/api/queryClient'
import { ApiError } from '@/api/ApiError'
export const POLL_INTERVAL_MS = 1000
export const LAZY_POLL_INTERVAL_MS = 10_000
export const POLL_INTERVAL_MS = 4_000
export const LAZY_POLL_INTERVAL_MS = 15_000
export const LobbyProvider = () => {
const room = useRoomContext()
@@ -8,7 +8,7 @@ import {
} from '../api/requestEntry'
export const WAIT_TIMEOUT_MS = 600000 // 10 minutes
export const POLL_INTERVAL_MS = 1000
export const POLL_INTERVAL_MS = 3_000
export const useLobby = ({
roomId,
@@ -2,7 +2,7 @@ import { useTranslation } from 'react-i18next'
import { RiHand } from '@remixicon/react'
import { ToggleButton } from '@/primitives'
import { css } from '@/styled-system/css'
import { useRoomContext } from '@livekit/components-react'
import { useIsSpeaking, useRoomContext } from '@livekit/components-react'
import { useRaisedHand } from '@/features/rooms/livekit/hooks/useRaisedHand'
import { useEffect, useRef, useState } from 'react'
import {
@@ -25,11 +25,11 @@ export const HandToggle = ({
const { t } = useTranslation('rooms', { keyPrefix: 'controls.hand' })
const room = useRoomContext()
const { isHandRaised, toggleRaisedHand } = useRaisedHand({
const { isHandRaised, toggleRaisedHand, lowerHand } = useRaisedHand({
participant: room.localParticipant,
})
const isSpeaking = room.localParticipant.isSpeaking
const isSpeaking = useIsSpeaking(room.localParticipant)
const speakingTimerRef = useRef<ReturnType<typeof setTimeout> | null>(null)
const [hasShownToast, setHasShownToast] = useState(false)
@@ -57,9 +57,10 @@ export const HandToggle = ({
if (shouldShowToast && !speakingTimerRef.current) {
speakingTimerRef.current = setTimeout(() => {
speakingTimerRef.current = null
setHasShownToast(true)
const onClose = () => {
if (isHandRaised) toggleRaisedHand()
lowerHand()
resetToastState()
}
showLowerHandToast(room.localParticipant, onClose)
@@ -70,7 +71,17 @@ export const HandToggle = ({
speakingTimerRef.current = null
}
// eslint-disable-next-line react-hooks/exhaustive-deps
}, [isSpeaking, isHandRaised, hasShownToast, toggleRaisedHand])
}, [isSpeaking, isHandRaised, hasShownToast, lowerHand])
// Clear any pending timer on unmount
useEffect(() => {
return () => {
if (speakingTimerRef.current) {
clearTimeout(speakingTimerRef.current)
speakingTimerRef.current = null
}
}
}, [])
const tooltipLabel = isHandRaised ? 'lower' : 'raise'
@@ -86,5 +86,16 @@ export function useRaisedHand({ participant }: useRaisedHandProps) {
}
}
return { isHandRaised, toggleRaisedHand }
const lowerHand = async () => {
if (!isLocal(participant)) return
try {
await raiseHand(false)
} catch (e) {
reportError('generic_failure', e, {
context: 'lower_raised_hand',
})
}
}
return { isHandRaised, toggleRaisedHand, lowerHand }
}
+7 -15
View File
@@ -35,30 +35,21 @@ const LoginHint = () => {
<div
className={css({
position: 'absolute',
top: '103px',
right: '110px',
top: 'calc(100% + 12px)',
right: 0,
zIndex: '100',
outline: 'none',
padding: '1.25rem',
maxWidth: '350px',
width: 'max-content',
maxWidth: 'min(350px, calc(100vw - 2rem))',
boxShadow: '0 2px 5px rgba(0 0 0 / 0.1)',
borderRadius: '1rem',
backgroundColor: 'primary.200',
display: 'none',
xsm: {
display: 'block',
},
sm: {
top: '131px',
right: '100px',
zIndex: '100',
},
_after: {
content: '""',
position: 'absolute',
top: '-10px',
right: '20%',
marginLeft: '-10px',
right: '1.5rem',
borderWidth: '0 10px 10px 10px',
borderStyle: 'solid',
borderColor: 'transparent transparent #E3E3FB transparent',
@@ -171,12 +162,13 @@ export const Header = () => {
<>
<div
className={css({
position: 'relative',
display: { base: 'none', xsm: 'block' },
})}
>
<LoginButton proConnectHint={false} />
</div>
<LoginHint />
</div>
</>
)}
{!!user && (
+4
View File
@@ -6,6 +6,10 @@ body,
height: 100%;
}
:root {
--avatar-cap-height: 0.7;
}
html.font-lexend {
--fonts-sans: 'Lexend Variable', ui-sans-serif, system-ui, sans-serif;
}