mirror of
https://github.com/1f349/dendrite.git
synced 2024-11-24 12:41:34 +00:00
f12982472c
Try to mitigate some issues with `/messages`
1069 lines
38 KiB
Go
1069 lines
38 KiB
Go
package storage_test
|
|
|
|
import (
|
|
"bytes"
|
|
"context"
|
|
"encoding/json"
|
|
"fmt"
|
|
"math"
|
|
"reflect"
|
|
"testing"
|
|
|
|
"github.com/matrix-org/dendrite/internal/sqlutil"
|
|
"github.com/matrix-org/dendrite/roomserver/api"
|
|
rstypes "github.com/matrix-org/dendrite/roomserver/types"
|
|
"github.com/matrix-org/dendrite/setup/config"
|
|
"github.com/matrix-org/dendrite/syncapi/storage"
|
|
"github.com/matrix-org/dendrite/syncapi/synctypes"
|
|
"github.com/matrix-org/dendrite/syncapi/types"
|
|
"github.com/matrix-org/dendrite/test"
|
|
"github.com/matrix-org/gomatrixserverlib"
|
|
"github.com/matrix-org/gomatrixserverlib/spec"
|
|
"github.com/stretchr/testify/assert"
|
|
"github.com/tidwall/gjson"
|
|
)
|
|
|
|
var ctx = context.Background()
|
|
|
|
func MustCreateDatabase(t *testing.T, dbType test.DBType) (storage.Database, func()) {
|
|
connStr, close := test.PrepareDBConnectionString(t, dbType)
|
|
cm := sqlutil.NewConnectionManager(nil, config.DatabaseOptions{})
|
|
db, err := storage.NewSyncServerDatasource(context.Background(), cm, &config.DatabaseOptions{
|
|
ConnectionString: config.DataSource(connStr),
|
|
})
|
|
if err != nil {
|
|
t.Fatalf("NewSyncServerDatasource returned %s", err)
|
|
}
|
|
return db, close
|
|
}
|
|
|
|
func MustWriteEvents(t *testing.T, db storage.Database, events []*rstypes.HeaderedEvent) (positions []types.StreamPosition) {
|
|
for _, ev := range events {
|
|
var addStateEvents []*rstypes.HeaderedEvent
|
|
var addStateEventIDs []string
|
|
var removeStateEventIDs []string
|
|
if ev.StateKey() != nil {
|
|
ev.StateKeyResolved = ev.StateKey()
|
|
addStateEvents = append(addStateEvents, ev)
|
|
addStateEventIDs = append(addStateEventIDs, ev.EventID())
|
|
}
|
|
pos, err := db.WriteEvent(ctx, ev, addStateEvents, addStateEventIDs, removeStateEventIDs, nil, false, gomatrixserverlib.HistoryVisibilityShared)
|
|
if err != nil {
|
|
t.Fatalf("WriteEvent failed: %s", err)
|
|
}
|
|
t.Logf("Event ID %s spos=%v depth=%v", ev.EventID(), pos, ev.Depth())
|
|
positions = append(positions, pos)
|
|
}
|
|
return
|
|
}
|
|
|
|
func TestWriteEvents(t *testing.T) {
|
|
test.WithAllDatabases(t, func(t *testing.T, dbType test.DBType) {
|
|
alice := test.NewUser(t)
|
|
r := test.NewRoom(t, alice)
|
|
db, close := MustCreateDatabase(t, dbType)
|
|
defer close()
|
|
MustWriteEvents(t, db, r.Events())
|
|
})
|
|
}
|
|
|
|
func WithSnapshot(t *testing.T, db storage.Database, f func(snapshot storage.DatabaseTransaction)) {
|
|
snapshot, err := db.NewDatabaseSnapshot(ctx)
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
f(snapshot)
|
|
if err := snapshot.Rollback(); err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
}
|
|
|
|
// These tests assert basic functionality of RecentEvents for PDUs
|
|
func TestRecentEventsPDU(t *testing.T) {
|
|
test.WithAllDatabases(t, func(t *testing.T, dbType test.DBType) {
|
|
db, close := MustCreateDatabase(t, dbType)
|
|
defer close()
|
|
alice := test.NewUser(t)
|
|
// dummy room to make sure SQL queries are filtering on room ID
|
|
MustWriteEvents(t, db, test.NewRoom(t, alice).Events())
|
|
|
|
// actual test room
|
|
r := test.NewRoom(t, alice)
|
|
r.CreateAndInsert(t, alice, "m.room.message", map[string]interface{}{"body": "hi"})
|
|
events := r.Events()
|
|
positions := MustWriteEvents(t, db, events)
|
|
|
|
// dummy room to make sure SQL queries are filtering on room ID
|
|
MustWriteEvents(t, db, test.NewRoom(t, alice).Events())
|
|
|
|
var latest types.StreamPosition
|
|
WithSnapshot(t, db, func(snapshot storage.DatabaseTransaction) {
|
|
var err error
|
|
if latest, err = snapshot.MaxStreamPositionForPDUs(ctx); err != nil {
|
|
t.Fatal("failed to get MaxStreamPositionForPDUs: %w", err)
|
|
}
|
|
})
|
|
|
|
testCases := []struct {
|
|
Name string
|
|
From types.StreamPosition
|
|
To types.StreamPosition
|
|
Limit int
|
|
ReverseOrder bool
|
|
WantEvents []*rstypes.HeaderedEvent
|
|
WantLimited bool
|
|
}{
|
|
// The purpose of this test is to make sure that incremental syncs are including up to the latest events.
|
|
// It's a basic sanity test that sync works. It creates a streaming position that is on the penultimate event.
|
|
// It makes sure the response includes the final event.
|
|
{
|
|
Name: "penultimate",
|
|
From: positions[len(positions)-2], // pretend we are at the penultimate event
|
|
To: latest,
|
|
Limit: 100,
|
|
WantEvents: events[len(events)-1:],
|
|
WantLimited: false,
|
|
},
|
|
// The purpose of this test is to check that limits can be applied and work.
|
|
// This is critical for big rooms hence the test here.
|
|
{
|
|
Name: "limited",
|
|
From: 0,
|
|
To: latest,
|
|
Limit: 1,
|
|
WantEvents: events[len(events)-1:],
|
|
WantLimited: true,
|
|
},
|
|
// The purpose of this test is to check that we can return every event with a high
|
|
// enough limit
|
|
{
|
|
Name: "large limited",
|
|
From: 0,
|
|
To: latest,
|
|
Limit: 100,
|
|
WantEvents: events,
|
|
WantLimited: false,
|
|
},
|
|
// The purpose of this test is to check that we can return events in reverse order
|
|
{
|
|
Name: "reverse",
|
|
From: positions[len(positions)-3], // 2 events back
|
|
To: latest,
|
|
Limit: 100,
|
|
ReverseOrder: true,
|
|
WantEvents: test.Reversed(events[len(events)-2:]),
|
|
WantLimited: false,
|
|
},
|
|
}
|
|
|
|
for i := range testCases {
|
|
tc := testCases[i]
|
|
t.Run(tc.Name, func(st *testing.T) {
|
|
var filter synctypes.RoomEventFilter
|
|
var gotEvents map[string]types.RecentEvents
|
|
var limited bool
|
|
filter.Limit = tc.Limit
|
|
WithSnapshot(t, db, func(snapshot storage.DatabaseTransaction) {
|
|
var err error
|
|
gotEvents, err = snapshot.RecentEvents(ctx, []string{r.ID}, types.Range{
|
|
From: tc.From,
|
|
To: tc.To,
|
|
}, &filter, !tc.ReverseOrder, true)
|
|
if err != nil {
|
|
st.Fatalf("failed to do sync: %s", err)
|
|
}
|
|
})
|
|
streamEvents := gotEvents[r.ID]
|
|
limited = streamEvents.Limited
|
|
if limited != tc.WantLimited {
|
|
st.Errorf("got limited=%v want %v", limited, tc.WantLimited)
|
|
}
|
|
if len(streamEvents.Events) != len(tc.WantEvents) {
|
|
st.Errorf("got %d events, want %d", len(gotEvents), len(tc.WantEvents))
|
|
}
|
|
|
|
for j := range streamEvents.Events {
|
|
if !reflect.DeepEqual(streamEvents.Events[j].JSON(), tc.WantEvents[j].JSON()) {
|
|
st.Errorf("event %d got %s want %s", j, string(streamEvents.Events[j].JSON()), string(tc.WantEvents[j].JSON()))
|
|
}
|
|
}
|
|
})
|
|
}
|
|
})
|
|
}
|
|
|
|
// The purpose of this test is to ensure that backfill does indeed go backwards, using a topology token
|
|
func TestGetEventsInRangeWithTopologyToken(t *testing.T) {
|
|
test.WithAllDatabases(t, func(t *testing.T, dbType test.DBType) {
|
|
db, close := MustCreateDatabase(t, dbType)
|
|
defer close()
|
|
alice := test.NewUser(t)
|
|
r := test.NewRoom(t, alice)
|
|
for i := 0; i < 10; i++ {
|
|
r.CreateAndInsert(t, alice, "m.room.message", map[string]interface{}{"body": fmt.Sprintf("hi %d", i)})
|
|
}
|
|
events := r.Events()
|
|
_ = MustWriteEvents(t, db, events)
|
|
|
|
WithSnapshot(t, db, func(snapshot storage.DatabaseTransaction) {
|
|
from := types.TopologyToken{Depth: math.MaxInt64, PDUPosition: math.MaxInt64}
|
|
t.Logf("max topo pos = %+v", from)
|
|
// head towards the beginning of time
|
|
to := types.TopologyToken{}
|
|
|
|
// backpaginate 5 messages starting at the latest position.
|
|
filter := &synctypes.RoomEventFilter{Limit: 5}
|
|
paginatedEvents, start, end, err := snapshot.GetEventsInTopologicalRange(ctx, &from, &to, r.ID, filter, true)
|
|
if err != nil {
|
|
t.Fatalf("GetEventsInTopologicalRange returned an error: %s", err)
|
|
}
|
|
gots := snapshot.StreamEventsToEvents(context.Background(), nil, paginatedEvents, nil)
|
|
test.AssertEventsEqual(t, gots, test.Reversed(events[len(events)-5:]))
|
|
assert.Equal(t, types.TopologyToken{Depth: 15, PDUPosition: 15}, start)
|
|
assert.Equal(t, types.TopologyToken{Depth: 11, PDUPosition: 11}, end)
|
|
})
|
|
})
|
|
}
|
|
|
|
// The purpose of this test is to ensure that backfilling returns no start/end if a given filter removes
|
|
// all events.
|
|
func TestGetEventsInRangeWithTopologyTokenNoEventsForFilter(t *testing.T) {
|
|
test.WithAllDatabases(t, func(t *testing.T, dbType test.DBType) {
|
|
db, close := MustCreateDatabase(t, dbType)
|
|
defer close()
|
|
alice := test.NewUser(t)
|
|
r := test.NewRoom(t, alice)
|
|
for i := 0; i < 10; i++ {
|
|
r.CreateAndInsert(t, alice, "m.room.message", map[string]interface{}{"body": fmt.Sprintf("hi %d", i)})
|
|
}
|
|
events := r.Events()
|
|
_ = MustWriteEvents(t, db, events)
|
|
|
|
WithSnapshot(t, db, func(snapshot storage.DatabaseTransaction) {
|
|
from := types.TopologyToken{Depth: math.MaxInt64, PDUPosition: math.MaxInt64}
|
|
t.Logf("max topo pos = %+v", from)
|
|
// head towards the beginning of time
|
|
to := types.TopologyToken{}
|
|
|
|
// backpaginate 20 messages starting at the latest position.
|
|
notTypes := []string{spec.MRoomRedaction}
|
|
senders := []string{alice.ID}
|
|
filter := &synctypes.RoomEventFilter{Limit: 20, NotTypes: ¬Types, Senders: &senders}
|
|
paginatedEvents, start, end, err := snapshot.GetEventsInTopologicalRange(ctx, &from, &to, r.ID, filter, true)
|
|
assert.NoError(t, err)
|
|
assert.Equal(t, 0, len(paginatedEvents))
|
|
// Even if we didn't get anything back due to the filter, we should still have start/end
|
|
assert.Equal(t, types.TopologyToken{Depth: 15, PDUPosition: 15}, start)
|
|
assert.Equal(t, types.TopologyToken{Depth: 1, PDUPosition: 1}, end)
|
|
})
|
|
})
|
|
}
|
|
|
|
func TestStreamToTopologicalPosition(t *testing.T) {
|
|
alice := test.NewUser(t)
|
|
r := test.NewRoom(t, alice)
|
|
|
|
testCases := []struct {
|
|
name string
|
|
roomID string
|
|
streamPos types.StreamPosition
|
|
backwardOrdering bool
|
|
wantToken types.TopologyToken
|
|
}{
|
|
{
|
|
name: "forward ordering found streamPos returns found position",
|
|
roomID: r.ID,
|
|
streamPos: 1,
|
|
backwardOrdering: false,
|
|
wantToken: types.TopologyToken{Depth: 1, PDUPosition: 1},
|
|
},
|
|
{
|
|
name: "forward ordering not found streamPos returns max position",
|
|
roomID: r.ID,
|
|
streamPos: 100,
|
|
backwardOrdering: false,
|
|
wantToken: types.TopologyToken{Depth: math.MaxInt64, PDUPosition: math.MaxInt64},
|
|
},
|
|
{
|
|
name: "backward ordering found streamPos returns found position",
|
|
roomID: r.ID,
|
|
streamPos: 1,
|
|
backwardOrdering: true,
|
|
wantToken: types.TopologyToken{Depth: 1, PDUPosition: 1},
|
|
},
|
|
{
|
|
name: "backward ordering not found streamPos returns maxDepth with param pduPosition",
|
|
roomID: r.ID,
|
|
streamPos: 100,
|
|
backwardOrdering: true,
|
|
wantToken: types.TopologyToken{Depth: 5, PDUPosition: 100},
|
|
},
|
|
{
|
|
name: "backward non-existent room returns zero token",
|
|
roomID: "!doesnotexist:localhost",
|
|
streamPos: 1,
|
|
backwardOrdering: true,
|
|
wantToken: types.TopologyToken{Depth: 0, PDUPosition: 1},
|
|
},
|
|
{
|
|
name: "forward non-existent room returns max token",
|
|
roomID: "!doesnotexist:localhost",
|
|
streamPos: 1,
|
|
backwardOrdering: false,
|
|
wantToken: types.TopologyToken{Depth: math.MaxInt64, PDUPosition: math.MaxInt64},
|
|
},
|
|
}
|
|
|
|
test.WithAllDatabases(t, func(t *testing.T, dbType test.DBType) {
|
|
db, close := MustCreateDatabase(t, dbType)
|
|
defer close()
|
|
|
|
txn, err := db.NewDatabaseTransaction(ctx)
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
defer txn.Rollback()
|
|
MustWriteEvents(t, db, r.Events())
|
|
|
|
for _, tc := range testCases {
|
|
t.Run(tc.name, func(t *testing.T) {
|
|
token, err := txn.StreamToTopologicalPosition(ctx, tc.roomID, tc.streamPos, tc.backwardOrdering)
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
if tc.wantToken != token {
|
|
t.Fatalf("expected token %q, got %q", tc.wantToken, token)
|
|
}
|
|
})
|
|
}
|
|
|
|
})
|
|
}
|
|
|
|
/*
|
|
// The purpose of this test is to make sure that backpagination returns all events, even if some events have the same depth.
|
|
// For cases where events have the same depth, the streaming token should be used to tie break so events written via WriteEvent
|
|
// will appear FIRST when going backwards. This test creates a DAG like:
|
|
// .-----> Message ---.
|
|
// Create -> Membership --------> Message -------> Message
|
|
// `-----> Message ---`
|
|
// depth 1 2 3 4
|
|
//
|
|
// With a total depth of 4. It tests that:
|
|
// - Backpagination over the whole fork should include all messages and not leave any out.
|
|
// - Backpagination from the middle of the fork should not return duplicates (things later than the token).
|
|
func TestGetEventsInRangeWithEventsSameDepth(t *testing.T) {
|
|
t.Parallel()
|
|
db := MustCreateDatabase(t)
|
|
|
|
var events []*types.HeaderedEvent
|
|
events = append(events, MustCreateEvent(t, testRoomID, nil, &gomatrixserverlib.EventBuilder{
|
|
Content: []byte(fmt.Sprintf(`{"room_version":"4","creator":"%s"}`, testUserIDA)),
|
|
Type: "m.room.create",
|
|
StateKey: &emptyStateKey,
|
|
Sender: testUserIDA,
|
|
Depth: int64(len(events) + 1),
|
|
}))
|
|
events = append(events, MustCreateEvent(t, testRoomID, []*types.HeaderedEvent{events[len(events)-1]}, &gomatrixserverlib.EventBuilder{
|
|
Content: []byte(`{"membership":"join"}`),
|
|
Type: "m.room.member",
|
|
StateKey: &testUserIDA,
|
|
Sender: testUserIDA,
|
|
Depth: int64(len(events) + 1),
|
|
}))
|
|
// fork the dag into three, same prev_events and depth
|
|
parent := []*types.HeaderedEvent{events[len(events)-1]}
|
|
depth := int64(len(events) + 1)
|
|
for i := 0; i < 3; i++ {
|
|
events = append(events, MustCreateEvent(t, testRoomID, parent, &gomatrixserverlib.EventBuilder{
|
|
Content: []byte(fmt.Sprintf(`{"body":"Message A %d"}`, i+1)),
|
|
Type: "m.room.message",
|
|
Sender: testUserIDA,
|
|
Depth: depth,
|
|
}))
|
|
}
|
|
// merge the fork, prev_events are all 3 messages, depth is increased by 1.
|
|
events = append(events, MustCreateEvent(t, testRoomID, events[len(events)-3:], &gomatrixserverlib.EventBuilder{
|
|
Content: []byte(`{"body":"Message merge"}`),
|
|
Type: "m.room.message",
|
|
Sender: testUserIDA,
|
|
Depth: depth + 1,
|
|
}))
|
|
MustWriteEvents(t, db, events)
|
|
fromLatest, err := db.EventPositionInTopology(ctx, events[len(events)-1].EventID())
|
|
if err != nil {
|
|
t.Fatalf("failed to get EventPositionInTopology: %s", err)
|
|
}
|
|
fromFork, err := db.EventPositionInTopology(ctx, events[len(events)-3].EventID()) // Message 2
|
|
if err != nil {
|
|
t.Fatalf("failed to get EventPositionInTopology for event: %s", err)
|
|
}
|
|
// head towards the beginning of time
|
|
to := types.TopologyToken{}
|
|
|
|
testCases := []struct {
|
|
Name string
|
|
From types.TopologyToken
|
|
Limit int
|
|
Wants []*types.HeaderedEvent
|
|
}{
|
|
{
|
|
Name: "Pagination over the whole fork",
|
|
From: fromLatest,
|
|
Limit: 5,
|
|
Wants: reversed(events[len(events)-5:]),
|
|
},
|
|
{
|
|
Name: "Paginating to the middle of the fork",
|
|
From: fromLatest,
|
|
Limit: 2,
|
|
Wants: reversed(events[len(events)-2:]),
|
|
},
|
|
{
|
|
Name: "Pagination FROM the middle of the fork",
|
|
From: fromFork,
|
|
Limit: 3,
|
|
Wants: reversed(events[len(events)-5 : len(events)-2]),
|
|
},
|
|
}
|
|
|
|
for _, tc := range testCases {
|
|
// backpaginate messages starting at the latest position.
|
|
paginatedEvents, err := db.GetEventsInTopologicalRange(ctx, &tc.From, &to, testRoomID, tc.Limit, true)
|
|
if err != nil {
|
|
t.Fatalf("%s GetEventsInRange returned an error: %s", tc.Name, err)
|
|
}
|
|
gots := gomatrixserverlib.HeaderedToClientEvents(db.StreamEventsToEvents(&testUserDeviceA, paginatedEvents), gomatrixserverlib.FormatAll)
|
|
assertEventsEqual(t, tc.Name, true, gots, tc.Wants)
|
|
}
|
|
}
|
|
|
|
// The purpose of this test is to make sure that the query to pull out events is honouring the room ID correctly.
|
|
// It works by creating two rooms with the same events in them, then selecting events by topological range.
|
|
// Specifically, we know that events with the same depth but lower stream positions are selected, and it's possible
|
|
// that this check isn't using the room ID if the brackets are wrong in the SQL query.
|
|
func TestGetEventsInTopologicalRangeMultiRoom(t *testing.T) {
|
|
t.Parallel()
|
|
db := MustCreateDatabase(t)
|
|
|
|
makeEvents := func(roomID string) (events []*types.HeaderedEvent) {
|
|
events = append(events, MustCreateEvent(t, roomID, nil, &gomatrixserverlib.EventBuilder{
|
|
Content: []byte(fmt.Sprintf(`{"room_version":"4","creator":"%s"}`, testUserIDA)),
|
|
Type: "m.room.create",
|
|
StateKey: &emptyStateKey,
|
|
Sender: testUserIDA,
|
|
Depth: int64(len(events) + 1),
|
|
}))
|
|
events = append(events, MustCreateEvent(t, roomID, []*types.HeaderedEvent{events[len(events)-1]}, &gomatrixserverlib.EventBuilder{
|
|
Content: []byte(`{"membership":"join"}`),
|
|
Type: "m.room.member",
|
|
StateKey: &testUserIDA,
|
|
Sender: testUserIDA,
|
|
Depth: int64(len(events) + 1),
|
|
}))
|
|
return
|
|
}
|
|
|
|
roomA := "!room_a:" + string(testOrigin)
|
|
roomB := "!room_b:" + string(testOrigin)
|
|
eventsA := makeEvents(roomA)
|
|
eventsB := makeEvents(roomB)
|
|
MustWriteEvents(t, db, eventsA)
|
|
MustWriteEvents(t, db, eventsB)
|
|
from, err := db.MaxTopologicalPosition(ctx, roomB)
|
|
if err != nil {
|
|
t.Fatalf("failed to get MaxTopologicalPosition: %s", err)
|
|
}
|
|
// head towards the beginning of time
|
|
to := types.TopologyToken{}
|
|
|
|
// Query using room B as room A was inserted first and hence A will have lower stream positions but identical depths,
|
|
// allowing this bug to surface.
|
|
paginatedEvents, err := db.GetEventsInTopologicalRange(ctx, &from, &to, roomB, 5, true)
|
|
if err != nil {
|
|
t.Fatalf("GetEventsInRange returned an error: %s", err)
|
|
}
|
|
gots := gomatrixserverlib.HeaderedToClientEvents(db.StreamEventsToEvents(&testUserDeviceA, paginatedEvents), gomatrixserverlib.FormatAll)
|
|
assertEventsEqual(t, "", true, gots, reversed(eventsB))
|
|
}
|
|
|
|
// The purpose of this test is to make sure that events are returned in the right *order* when they have been inserted in a manner similar to
|
|
// how any kind of backfill operation will insert the events. This test inserts the SimpleRoom events in a manner similar to how backfill over
|
|
// federation would:
|
|
// - First inserts join event of test user C
|
|
// - Inserts chunks of history in strata e.g (25-30, 20-25, 15-20, 10-15, 5-10, 0-5).
|
|
// The test then does a backfill to ensure that the response is ordered correctly according to depth.
|
|
func TestGetEventsInRangeWithEventsInsertedLikeBackfill(t *testing.T) {
|
|
t.Parallel()
|
|
db := MustCreateDatabase(t)
|
|
events, _ := SimpleRoom(t, testRoomID, testUserIDA, testUserIDB)
|
|
|
|
// "federation" join
|
|
userC := fmt.Sprintf("@radiance:%s", testOrigin)
|
|
joinEvent := MustCreateEvent(t, testRoomID, []*types.HeaderedEvent{events[len(events)-1]}, &gomatrixserverlib.EventBuilder{
|
|
Content: []byte(`{"membership":"join"}`),
|
|
Type: "m.room.member",
|
|
StateKey: &userC,
|
|
Sender: userC,
|
|
Depth: int64(len(events) + 1),
|
|
})
|
|
MustWriteEvents(t, db, []*types.HeaderedEvent{joinEvent})
|
|
|
|
// Sync will return this for the prev_batch
|
|
from := topologyTokenBefore(t, db, joinEvent.EventID())
|
|
|
|
// inject events in batches as if they were from backfill
|
|
// e.g [1,2,3,4,5,6] => [4,5,6] , [1,2,3]
|
|
chunkSize := 5
|
|
for i := len(events); i >= 0; i -= chunkSize {
|
|
start := i - chunkSize
|
|
if start < 0 {
|
|
start = 0
|
|
}
|
|
backfill := events[start:i]
|
|
MustWriteEvents(t, db, backfill)
|
|
}
|
|
|
|
// head towards the beginning of time
|
|
to := types.TopologyToken{}
|
|
|
|
// starting at `from`, backpaginate to the beginning of time, asserting as we go.
|
|
chunkSize = 3
|
|
events = reversed(events)
|
|
for i := 0; i < len(events); i += chunkSize {
|
|
paginatedEvents, err := db.GetEventsInTopologicalRange(ctx, from, &to, testRoomID, chunkSize, true)
|
|
if err != nil {
|
|
t.Fatalf("GetEventsInRange returned an error: %s", err)
|
|
}
|
|
gots := gomatrixserverlib.HeaderedToClientEvents(db.StreamEventsToEvents(&testUserDeviceA, paginatedEvents), gomatrixserverlib.FormatAll)
|
|
endi := i + chunkSize
|
|
if endi > len(events) {
|
|
endi = len(events)
|
|
}
|
|
assertEventsEqual(t, from.String(), true, gots, events[i:endi])
|
|
from = topologyTokenBefore(t, db, paginatedEvents[len(paginatedEvents)-1].EventID())
|
|
}
|
|
}
|
|
*/
|
|
|
|
func TestSendToDeviceBehaviour(t *testing.T) {
|
|
t.Parallel()
|
|
alice := test.NewUser(t)
|
|
bob := test.NewUser(t)
|
|
deviceID := "one"
|
|
test.WithAllDatabases(t, func(t *testing.T, dbType test.DBType) {
|
|
db, close := MustCreateDatabase(t, dbType)
|
|
defer close()
|
|
// At this point there should be no messages. We haven't sent anything
|
|
// yet.
|
|
|
|
WithSnapshot(t, db, func(snapshot storage.DatabaseTransaction) {
|
|
_, events, err := snapshot.SendToDeviceUpdatesForSync(ctx, alice.ID, deviceID, 0, 100)
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
if len(events) != 0 {
|
|
t.Fatal("first call should have no updates")
|
|
}
|
|
})
|
|
|
|
// Try sending a message.
|
|
streamPos, err := db.StoreNewSendForDeviceMessage(ctx, alice.ID, deviceID, gomatrixserverlib.SendToDeviceEvent{
|
|
Sender: bob.ID,
|
|
Type: "m.type",
|
|
Content: json.RawMessage("{}"),
|
|
})
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
|
|
WithSnapshot(t, db, func(snapshot storage.DatabaseTransaction) {
|
|
// At this point we should get exactly one message. We're sending the sync position
|
|
// that we were given from the update and the send-to-device update will be updated
|
|
// in the database to reflect that this was the sync position we sent the message at.
|
|
var events []types.SendToDeviceEvent
|
|
streamPos, events, err = snapshot.SendToDeviceUpdatesForSync(ctx, alice.ID, deviceID, 0, streamPos)
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
if count := len(events); count != 1 {
|
|
t.Fatalf("second call should have one update, got %d", count)
|
|
}
|
|
|
|
// At this point we should still have one message because we haven't progressed the
|
|
// sync position yet. This is equivalent to the client failing to /sync and retrying
|
|
// with the same position.
|
|
streamPos, events, err = snapshot.SendToDeviceUpdatesForSync(ctx, alice.ID, deviceID, 0, streamPos)
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
if len(events) != 1 {
|
|
t.Fatal("third call should have one update still")
|
|
}
|
|
})
|
|
|
|
err = db.CleanSendToDeviceUpdates(context.Background(), alice.ID, deviceID, streamPos)
|
|
if err != nil {
|
|
return
|
|
}
|
|
|
|
WithSnapshot(t, db, func(snapshot storage.DatabaseTransaction) {
|
|
// At this point we should now have no updates, because we've progressed the sync
|
|
// position. Therefore the update from before will not be sent again.
|
|
var events []types.SendToDeviceEvent
|
|
_, events, err = snapshot.SendToDeviceUpdatesForSync(ctx, alice.ID, deviceID, streamPos, streamPos+10)
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
if len(events) != 0 {
|
|
t.Fatal("fourth call should have no updates")
|
|
}
|
|
|
|
// At this point we should still have no updates, because no new updates have been
|
|
// sent.
|
|
_, events, err = snapshot.SendToDeviceUpdatesForSync(ctx, alice.ID, deviceID, streamPos, streamPos+10)
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
if len(events) != 0 {
|
|
t.Fatal("fifth call should have no updates")
|
|
}
|
|
})
|
|
|
|
// Send some more messages and verify the ordering is correct ("in order of arrival")
|
|
var lastPos types.StreamPosition = 0
|
|
for i := 0; i < 10; i++ {
|
|
streamPos, err = db.StoreNewSendForDeviceMessage(ctx, alice.ID, deviceID, gomatrixserverlib.SendToDeviceEvent{
|
|
Sender: bob.ID,
|
|
Type: "m.type",
|
|
Content: json.RawMessage(fmt.Sprintf(`{"count":%d}`, i)),
|
|
})
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
lastPos = streamPos
|
|
}
|
|
|
|
WithSnapshot(t, db, func(snapshot storage.DatabaseTransaction) {
|
|
_, events, err := snapshot.SendToDeviceUpdatesForSync(ctx, alice.ID, deviceID, 0, lastPos)
|
|
if err != nil {
|
|
t.Fatalf("unable to get events: %v", err)
|
|
}
|
|
|
|
for i := 0; i < 10; i++ {
|
|
want := json.RawMessage(fmt.Sprintf(`{"count":%d}`, i))
|
|
got := events[i].Content
|
|
if !bytes.Equal(got, want) {
|
|
t.Fatalf("messages are out of order\nwant: %s\ngot: %s", string(want), string(got))
|
|
}
|
|
}
|
|
})
|
|
})
|
|
}
|
|
|
|
/*
|
|
func TestInviteBehaviour(t *testing.T) {
|
|
db := MustCreateDatabase(t)
|
|
inviteRoom1 := "!inviteRoom1:somewhere"
|
|
inviteEvent1 := MustCreateEvent(t, inviteRoom1, nil, &gomatrixserverlib.EventBuilder{
|
|
Content: []byte(`{"membership":"invite"}`),
|
|
Type: "m.room.member",
|
|
StateKey: &testUserIDA,
|
|
Sender: "@inviteUser1:somewhere",
|
|
})
|
|
inviteRoom2 := "!inviteRoom2:somewhere"
|
|
inviteEvent2 := MustCreateEvent(t, inviteRoom2, nil, &gomatrixserverlib.EventBuilder{
|
|
Content: []byte(`{"membership":"invite"}`),
|
|
Type: "m.room.member",
|
|
StateKey: &testUserIDA,
|
|
Sender: "@inviteUser2:somewhere",
|
|
})
|
|
for _, ev := range []*types.HeaderedEvent{inviteEvent1, inviteEvent2} {
|
|
_, err := db.AddInviteEvent(ctx, ev)
|
|
if err != nil {
|
|
t.Fatalf("Failed to AddInviteEvent: %s", err)
|
|
}
|
|
}
|
|
latest, err := db.SyncPosition(ctx)
|
|
if err != nil {
|
|
t.Fatalf("failed to get SyncPosition: %s", err)
|
|
}
|
|
// both invite events should appear in a new sync
|
|
beforeRetireRes := types.NewResponse()
|
|
beforeRetireRes, err = db.IncrementalSync(ctx, beforeRetireRes, testUserDeviceA, types.StreamingToken{}, latest, 0, false)
|
|
if err != nil {
|
|
t.Fatalf("IncrementalSync failed: %s", err)
|
|
}
|
|
assertInvitedToRooms(t, beforeRetireRes, []string{inviteRoom1, inviteRoom2})
|
|
|
|
// retire one event: a fresh sync should just return 1 invite room
|
|
if _, err = db.RetireInviteEvent(ctx, inviteEvent1.EventID()); err != nil {
|
|
t.Fatalf("Failed to RetireInviteEvent: %s", err)
|
|
}
|
|
latest, err = db.SyncPosition(ctx)
|
|
if err != nil {
|
|
t.Fatalf("failed to get SyncPosition: %s", err)
|
|
}
|
|
res := types.NewResponse()
|
|
res, err = db.IncrementalSync(ctx, res, testUserDeviceA, types.StreamingToken{}, latest, 0, false)
|
|
if err != nil {
|
|
t.Fatalf("IncrementalSync failed: %s", err)
|
|
}
|
|
assertInvitedToRooms(t, res, []string{inviteRoom2})
|
|
|
|
// a sync after we have received both invites should result in a leave for the retired room
|
|
res = types.NewResponse()
|
|
res, err = db.IncrementalSync(ctx, res, testUserDeviceA, beforeRetireRes.NextBatch, latest, 0, false)
|
|
if err != nil {
|
|
t.Fatalf("IncrementalSync failed: %s", err)
|
|
}
|
|
assertInvitedToRooms(t, res, []string{})
|
|
if _, ok := res.Rooms.Leave[inviteRoom1]; !ok {
|
|
t.Fatalf("IncrementalSync: expected to see room left after it was retired but it wasn't")
|
|
}
|
|
}
|
|
|
|
func assertInvitedToRooms(t *testing.T, res *types.Response, roomIDs []string) {
|
|
t.Helper()
|
|
if len(res.Rooms.Invite) != len(roomIDs) {
|
|
t.Fatalf("got %d invited rooms, want %d", len(res.Rooms.Invite), len(roomIDs))
|
|
}
|
|
for _, roomID := range roomIDs {
|
|
if _, ok := res.Rooms.Invite[roomID]; !ok {
|
|
t.Fatalf("missing room ID %s", roomID)
|
|
}
|
|
}
|
|
}
|
|
|
|
func assertEventsEqual(t *testing.T, msg string, checkRoomID bool, gots []gomatrixserverlib.ClientEvent, wants []*types.HeaderedEvent) {
|
|
t.Helper()
|
|
if len(gots) != len(wants) {
|
|
t.Fatalf("%s response returned %d events, want %d", msg, len(gots), len(wants))
|
|
}
|
|
for i := range gots {
|
|
g := gots[i]
|
|
w := wants[i]
|
|
if g.EventID != w.EventID() {
|
|
t.Errorf("%s event[%d] event_id mismatch: got %s want %s", msg, i, g.EventID, w.EventID())
|
|
}
|
|
if g.Sender != w.Sender() {
|
|
t.Errorf("%s event[%d] sender mismatch: got %s want %s", msg, i, g.Sender, w.Sender())
|
|
}
|
|
if checkRoomID && g.RoomID != w.RoomID() {
|
|
t.Errorf("%s event[%d] room_id mismatch: got %s want %s", msg, i, g.RoomID, w.RoomID())
|
|
}
|
|
if g.Type != w.Type() {
|
|
t.Errorf("%s event[%d] event type mismatch: got %s want %s", msg, i, g.Type, w.Type())
|
|
}
|
|
if g.OriginServerTS != w.OriginServerTS() {
|
|
t.Errorf("%s event[%d] origin_server_ts mismatch: got %v want %v", msg, i, g.OriginServerTS, w.OriginServerTS())
|
|
}
|
|
if string(g.Content) != string(w.Content()) {
|
|
t.Errorf("%s event[%d] content mismatch: got %s want %s", msg, i, string(g.Content), string(w.Content()))
|
|
}
|
|
if string(g.Unsigned) != string(w.Unsigned()) {
|
|
t.Errorf("%s event[%d] unsigned mismatch: got %s want %s", msg, i, string(g.Unsigned), string(w.Unsigned()))
|
|
}
|
|
if (g.StateKey == nil && w.StateKey() != nil) || (g.StateKey != nil && w.StateKey() == nil) {
|
|
t.Errorf("%s event[%d] state_key [not] missing: got %v want %v", msg, i, g.StateKey, w.StateKey())
|
|
continue
|
|
}
|
|
if g.StateKey != nil {
|
|
if !w.StateKeyEquals(*g.StateKey) {
|
|
t.Errorf("%s event[%d] state_key mismatch: got %s want %s", msg, i, *g.StateKey, *w.StateKey())
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
func topologyTokenBefore(t *testing.T, db storage.Database, eventID string) *types.TopologyToken {
|
|
tok, err := db.EventPositionInTopology(ctx, eventID)
|
|
if err != nil {
|
|
t.Fatalf("failed to get EventPositionInTopology: %s", err)
|
|
}
|
|
tok.Decrement()
|
|
return &tok
|
|
}
|
|
*/
|
|
|
|
func pointer[t any](s t) *t {
|
|
return &s
|
|
}
|
|
|
|
func TestRoomSummary(t *testing.T) {
|
|
|
|
alice := test.NewUser(t)
|
|
bob := test.NewUser(t)
|
|
charlie := test.NewUser(t)
|
|
|
|
// Create some dummy users
|
|
moreUsers := []*test.User{}
|
|
moreUserIDs := []string{}
|
|
for i := 0; i < 10; i++ {
|
|
u := test.NewUser(t)
|
|
moreUsers = append(moreUsers, u)
|
|
moreUserIDs = append(moreUserIDs, u.ID)
|
|
}
|
|
|
|
testCases := []struct {
|
|
name string
|
|
wantSummary *types.Summary
|
|
additionalEvents func(t *testing.T, room *test.Room)
|
|
}{
|
|
{
|
|
name: "after initial creation",
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(1), InvitedMemberCount: pointer(0), Heroes: []string{}},
|
|
},
|
|
{
|
|
name: "invited user",
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(1), InvitedMemberCount: pointer(1), Heroes: []string{bob.ID}},
|
|
additionalEvents: func(t *testing.T, room *test.Room) {
|
|
room.CreateAndInsert(t, alice, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "invite",
|
|
}, test.WithStateKey(bob.ID))
|
|
},
|
|
},
|
|
{
|
|
name: "invited user, but declined",
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(1), InvitedMemberCount: pointer(0), Heroes: []string{bob.ID}},
|
|
additionalEvents: func(t *testing.T, room *test.Room) {
|
|
room.CreateAndInsert(t, alice, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "invite",
|
|
}, test.WithStateKey(bob.ID))
|
|
room.CreateAndInsert(t, bob, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "leave",
|
|
}, test.WithStateKey(bob.ID))
|
|
},
|
|
},
|
|
{
|
|
name: "joined user after invitation",
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(2), InvitedMemberCount: pointer(0), Heroes: []string{bob.ID}},
|
|
additionalEvents: func(t *testing.T, room *test.Room) {
|
|
room.CreateAndInsert(t, alice, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "invite",
|
|
}, test.WithStateKey(bob.ID))
|
|
room.CreateAndInsert(t, bob, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "join",
|
|
}, test.WithStateKey(bob.ID))
|
|
},
|
|
},
|
|
{
|
|
name: "multiple joined user",
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(3), InvitedMemberCount: pointer(0), Heroes: []string{charlie.ID, bob.ID}},
|
|
additionalEvents: func(t *testing.T, room *test.Room) {
|
|
room.CreateAndInsert(t, charlie, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "join",
|
|
}, test.WithStateKey(charlie.ID))
|
|
room.CreateAndInsert(t, bob, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "join",
|
|
}, test.WithStateKey(bob.ID))
|
|
},
|
|
},
|
|
{
|
|
name: "multiple joined/invited user",
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(2), InvitedMemberCount: pointer(1), Heroes: []string{charlie.ID, bob.ID}},
|
|
additionalEvents: func(t *testing.T, room *test.Room) {
|
|
room.CreateAndInsert(t, alice, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "invite",
|
|
}, test.WithStateKey(charlie.ID))
|
|
room.CreateAndInsert(t, bob, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "join",
|
|
}, test.WithStateKey(bob.ID))
|
|
},
|
|
},
|
|
{
|
|
name: "multiple joined/invited/left user",
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(1), InvitedMemberCount: pointer(1), Heroes: []string{charlie.ID}},
|
|
additionalEvents: func(t *testing.T, room *test.Room) {
|
|
room.CreateAndInsert(t, alice, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "invite",
|
|
}, test.WithStateKey(charlie.ID))
|
|
room.CreateAndInsert(t, bob, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "join",
|
|
}, test.WithStateKey(bob.ID))
|
|
room.CreateAndInsert(t, bob, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "leave",
|
|
}, test.WithStateKey(bob.ID))
|
|
},
|
|
},
|
|
{
|
|
name: "leaving user after joining",
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(1), InvitedMemberCount: pointer(0), Heroes: []string{bob.ID}},
|
|
additionalEvents: func(t *testing.T, room *test.Room) {
|
|
room.CreateAndInsert(t, bob, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "join",
|
|
}, test.WithStateKey(bob.ID))
|
|
room.CreateAndInsert(t, bob, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "leave",
|
|
}, test.WithStateKey(bob.ID))
|
|
},
|
|
},
|
|
{
|
|
name: "many users", // heroes ordered by stream id
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(len(moreUserIDs) + 1), InvitedMemberCount: pointer(0), Heroes: moreUserIDs[:5]},
|
|
additionalEvents: func(t *testing.T, room *test.Room) {
|
|
for _, x := range moreUsers {
|
|
room.CreateAndInsert(t, x, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "join",
|
|
}, test.WithStateKey(x.ID))
|
|
}
|
|
},
|
|
},
|
|
{
|
|
name: "canonical alias set",
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(2), InvitedMemberCount: pointer(0), Heroes: []string{}},
|
|
additionalEvents: func(t *testing.T, room *test.Room) {
|
|
room.CreateAndInsert(t, bob, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "join",
|
|
}, test.WithStateKey(bob.ID))
|
|
room.CreateAndInsert(t, alice, spec.MRoomCanonicalAlias, map[string]interface{}{
|
|
"alias": "myalias",
|
|
}, test.WithStateKey(""))
|
|
},
|
|
},
|
|
{
|
|
name: "room name set",
|
|
wantSummary: &types.Summary{JoinedMemberCount: pointer(2), InvitedMemberCount: pointer(0), Heroes: []string{}},
|
|
additionalEvents: func(t *testing.T, room *test.Room) {
|
|
room.CreateAndInsert(t, bob, spec.MRoomMember, map[string]interface{}{
|
|
"membership": "join",
|
|
}, test.WithStateKey(bob.ID))
|
|
room.CreateAndInsert(t, alice, spec.MRoomName, map[string]interface{}{
|
|
"name": "my room name",
|
|
}, test.WithStateKey(""))
|
|
},
|
|
},
|
|
}
|
|
|
|
test.WithAllDatabases(t, func(t *testing.T, dbType test.DBType) {
|
|
db, close := MustCreateDatabase(t, dbType)
|
|
defer close()
|
|
|
|
for _, tc := range testCases {
|
|
t.Run(tc.name, func(t *testing.T) {
|
|
|
|
r := test.NewRoom(t, alice)
|
|
|
|
if tc.additionalEvents != nil {
|
|
tc.additionalEvents(t, r)
|
|
}
|
|
|
|
// write the room before creating a transaction
|
|
MustWriteEvents(t, db, r.Events())
|
|
|
|
transaction, err := db.NewDatabaseTransaction(ctx)
|
|
assert.NoError(t, err)
|
|
defer transaction.Rollback()
|
|
|
|
summary, err := transaction.GetRoomSummary(ctx, r.ID, alice.ID)
|
|
assert.NoError(t, err)
|
|
assert.Equal(t, tc.wantSummary, summary)
|
|
})
|
|
}
|
|
})
|
|
}
|
|
|
|
func TestRecentEvents(t *testing.T) {
|
|
alice := test.NewUser(t)
|
|
room1 := test.NewRoom(t, alice)
|
|
room2 := test.NewRoom(t, alice)
|
|
roomIDs := []string{room1.ID, room2.ID}
|
|
rooms := map[string]*test.Room{
|
|
room1.ID: room1,
|
|
room2.ID: room2,
|
|
}
|
|
|
|
test.WithAllDatabases(t, func(t *testing.T, dbType test.DBType) {
|
|
filter := synctypes.DefaultRoomEventFilter()
|
|
db, close := MustCreateDatabase(t, dbType)
|
|
t.Cleanup(close)
|
|
|
|
MustWriteEvents(t, db, room1.Events())
|
|
MustWriteEvents(t, db, room2.Events())
|
|
|
|
transaction, err := db.NewDatabaseTransaction(ctx)
|
|
assert.NoError(t, err)
|
|
defer transaction.Rollback()
|
|
|
|
// get all recent events from 0 to 100 (we only created 5 events, so we should get 5 back)
|
|
roomEvs, err := transaction.RecentEvents(ctx, roomIDs, types.Range{From: 0, To: 100}, &filter, true, true)
|
|
assert.NoError(t, err)
|
|
assert.Equal(t, len(roomEvs), 2, "unexpected recent events response")
|
|
for _, recentEvents := range roomEvs {
|
|
assert.Equal(t, 5, len(recentEvents.Events), "unexpected recent events for room")
|
|
}
|
|
|
|
// update the filter to only return one event
|
|
filter.Limit = 1
|
|
roomEvs, err = transaction.RecentEvents(ctx, roomIDs, types.Range{From: 0, To: 100}, &filter, true, true)
|
|
assert.NoError(t, err)
|
|
assert.Equal(t, len(roomEvs), 2, "unexpected recent events response")
|
|
for roomID, recentEvents := range roomEvs {
|
|
origEvents := rooms[roomID].Events()
|
|
assert.Equal(t, true, recentEvents.Limited, "expected events to be limited")
|
|
assert.Equal(t, 1, len(recentEvents.Events), "unexpected recent events for room")
|
|
assert.Equal(t, origEvents[len(origEvents)-1].EventID(), recentEvents.Events[0].EventID())
|
|
}
|
|
|
|
// not chronologically ordered still returns the events in order (given ORDER BY id DESC)
|
|
roomEvs, err = transaction.RecentEvents(ctx, roomIDs, types.Range{From: 0, To: 100}, &filter, false, true)
|
|
assert.NoError(t, err)
|
|
assert.Equal(t, len(roomEvs), 2, "unexpected recent events response")
|
|
for roomID, recentEvents := range roomEvs {
|
|
origEvents := rooms[roomID].Events()
|
|
assert.Equal(t, true, recentEvents.Limited, "expected events to be limited")
|
|
assert.Equal(t, 1, len(recentEvents.Events), "unexpected recent events for room")
|
|
assert.Equal(t, origEvents[len(origEvents)-1].EventID(), recentEvents.Events[0].EventID())
|
|
}
|
|
})
|
|
}
|
|
|
|
type FakeQuerier struct {
|
|
api.QuerySenderIDAPI
|
|
}
|
|
|
|
func (f *FakeQuerier) QueryUserIDForSender(ctx context.Context, roomID spec.RoomID, senderID spec.SenderID) (*spec.UserID, error) {
|
|
return spec.NewUserID(string(senderID), true)
|
|
}
|
|
|
|
func TestRedaction(t *testing.T) {
|
|
alice := test.NewUser(t)
|
|
room := test.NewRoom(t, alice)
|
|
|
|
redactedEvent := room.CreateAndInsert(t, alice, "m.room.message", map[string]interface{}{"body": "hi"})
|
|
redactionEvent := room.CreateEvent(t, alice, spec.MRoomRedaction, map[string]string{"redacts": redactedEvent.EventID()})
|
|
test.WithAllDatabases(t, func(t *testing.T, dbType test.DBType) {
|
|
db, close := MustCreateDatabase(t, dbType)
|
|
t.Cleanup(close)
|
|
MustWriteEvents(t, db, room.Events())
|
|
|
|
err := db.RedactEvent(context.Background(), redactedEvent.EventID(), redactionEvent, &FakeQuerier{})
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
|
|
evs, err := db.Events(context.Background(), []string{redactedEvent.EventID()})
|
|
if err != nil {
|
|
t.Fatal(err)
|
|
}
|
|
|
|
if len(evs) != 1 {
|
|
t.Fatalf("expected 1 event, got %d", len(evs))
|
|
}
|
|
|
|
// check a few fields which shouldn't be there in unsigned
|
|
authEvs := gjson.GetBytes(evs[0].Unsigned(), "redacted_because.auth_events")
|
|
if authEvs.Exists() {
|
|
t.Error("unexpected auth_events in redacted event")
|
|
}
|
|
prevEvs := gjson.GetBytes(evs[0].Unsigned(), "redacted_because.prev_events")
|
|
if prevEvs.Exists() {
|
|
t.Error("unexpected auth_events in redacted event")
|
|
}
|
|
depth := gjson.GetBytes(evs[0].Unsigned(), "redacted_because.depth")
|
|
if depth.Exists() {
|
|
t.Error("unexpected auth_events in redacted event")
|
|
}
|
|
})
|
|
}
|