Implement /_matrix/client/v1/rooms/{roomId}/threads (I am being lazy so can someone finish off this for me?)
#3404
Conversation
|
this pr doesn't need tests because it's obviously working |
|
It needs tests so when new features are added or refactored, it won't break, and other devs who submit different pull requests, don't have to test this feature each time they submit a new PR. Tests are for future stability, the current feature is obviously always working, because if it wouldn't work, the dev would spend more time on it. But when 10 other features are added, nobody wants to manually test this feature 10 times to avoid breaking it while implementing those new features. |
Sure I will do it later |
|
@S7evinK can you run the workflow to test it? |
| tab, _, close := newRelationsTable(t, dbType) | ||
| defer close() | ||
|
|
||
| err = tab.InsertRelation(ctx, nil, room.ID, firstEvent.EventID(), threadReplyEvent.EventID(), "m.room.message", threadReplyEvent.EventID()) |
There was a problem hiding this comment.
Looks like the input is wrong:
InsertRelation(ctx context.Context, txn *sql.Tx, roomID, eventID, childEventID, childEventType, relType string) (err error)| t.Fatal(err) | ||
| } | ||
| var eventIds []string | ||
| eventIds, _, err = tab.SelectThreads(ctx, nil, room.ID, "", 0, 100) |
| t.Fatalf("eventID mismatch: got %s, want %s", eventID, expected) | ||
| } | ||
| } | ||
| eventIds, _, err = tab.SelectThreads(ctx, nil, room.ID, alice.ID, 0, 100) |
There was a problem hiding this comment.
Same as above, missing error check.
|
|
||
| var userID string | ||
| if include == "participated" { | ||
| _, err := spec.NewUserID(device.UserID, true) |
There was a problem hiding this comment.
| _, err := spec.NewUserID(device.UserID, true) | |
| _, err = spec.NewUserID(device.UserID, true) |
| } | ||
|
|
||
| for _, event := range headeredEvents { | ||
| ce, err := synctypes.ToClientEvent(event, synctypes.FormatAll, func(roomID spec.RoomID, senderID spec.SenderID) (*spec.UserID, error) { |
There was a problem hiding this comment.
I think the roomID is not needed, as we're already requesting for a specific room?
| ce, err := synctypes.ToClientEvent(event, synctypes.FormatAll, func(roomID spec.RoomID, senderID spec.SenderID) (*spec.UserID, error) { | |
| ce, err := synctypes.ToClientEvent(event, synctypes.FormatSync, func(roomID spec.RoomID, senderID spec.SenderID) (*spec.UserID, error) { |
| @@ -811,3 +811,39 @@ func (d *DatabaseTransaction) RelationsFor(ctx context.Context, roomID, eventID, | |||
|
|
|||
| return events, prevBatch, nextBatch, nil | |||
| } | |||
|
|
|||
| func (d *DatabaseTransaction) ThreadsFor(ctx context.Context, roomID, userID string, from types.StreamPosition, limit uint64) ( | |||
| events []*rstypes.HeaderedEvent, prevBatch, nextBatch string, err error, | |||
There was a problem hiding this comment.
prevBatch is always empty, so can be removed.
| room := test.NewRoom(t, alice) | ||
|
|
||
| firstEvent := room.CreateAndInsert(t, alice, "m.room.message", map[string]interface{}{ | ||
| "body": "first message", | ||
| }) | ||
| threadReplyEvent := room.CreateAndInsert(t, alice, "m.room.message", map[string]interface{}{ | ||
| "body": "thread reply", | ||
| "m.relates_to": map[string]interface{}{ | ||
| "event_id": firstEvent.EventID(), | ||
| "rel_type": threadRelType, | ||
| }, | ||
| }) |
There was a problem hiding this comment.
This is an in-memory room, so when you query the relation below, there won't be any events in the database. This means you'll also need to create a newOutputRoomEventsTable(t, dbType) below, like you did with the relations table. Then insert the events.
There was a problem hiding this comment.
This file needs to be formatted.
Co-authored-by: Till <2353100+S7evinK@users.noreply.github.com>
qwqtoday
changed the title
/_matrix/client/v1/rooms/{roomId}/threads/_matrix/client/v1/rooms/{roomId}/threads (I am being lazy so can someone finish off this for me?)
|
Thanks for contributing to Dendrite! The repo has been cloned into element-hq/dendrite and made the switch to AGPL as per https://matrix.org/blog/2023/11/06/future-of-synapse-dendrite/. If you still wish to contribute under the new license, feel free to create a new PR over at element-hq/dendrite. |
Signed-off-by:
Leung Ho Ching <lukas.leung@hotmail.com>Pull Request Checklist