Uh oh!
There was an error while loading. Please reload this page.
Migrate webhook-driven connectors onto scheduleDrain; extend the primitive - #265
Merged
Merged
Conversation
Extends the drain primitive with the two capabilities the migration
surfaced, then moves every notification-driven connector onto it:
SDK:
- scheduleDrain accepts handlerArgs (extra serializable args appended
after the ids slice) so per-scope drains — e.g. a per-channel key —
can carry the scope to the handler.
- The handler may return { retry: ids } to report PARTIAL failure: just
those ids stay pending with bumped attempt counters while the rest of
the slice is released.
Connectors:
- Gmail + combined Google (mail): signal-only scheduleDrain replaces
the hand-wired coalescing task; the history cursor remains the source
of work.
- MS Teams + Slack: per-channel signal-only drains with handlerArgs.
The sync window is now computed at drain time inside the handler, so
the webhook path no longer writes any state.
- Outlook Mail + combined Outlook (mail): the hand-rolled pending-key
machinery is replaced by id-mode scheduleDrain with the retry
contract (probe failures retry per-id; drafts/excluded/deleted ids
are released). A one-time upgrade() migration re-records pre-drain
pending ids and clears the old state; already-queued callbacks
re-record their ids through the same path.
Also fixes the Slack syncCustomEmoji test's incomplete tool mock
(missing scheduleRecurring), which failed independently of any change.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for freeto join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Follow-up to #263: makes
scheduleDrainthe real implementation everywhere, not just the documented pattern.SDK extensions (things the migration itself surfaced)
handlerArgs— extra serializable arguments appended after the ids slice (handler(ids, ...handlerArgs)), for per-scope drains such as a per-channel key whose handler needs the channel id.{ retry: ids }: just those ids stay pending with bumped attempt counters; the rest of the slice is released. Previously a handler could only fail the whole slice.Both additive; changeset included. 17 drain tests.
Migrations
handlerArgs)sync_statewrite per notification).retryupgrade()migration re-records pre-drain pending ids; already-queued callbacks re-record through the same path, so nothing in flight is lost across the deploy.Net: the four connectors that each hand-built coalescing/bounding/bookkeeping now share the platform implementation — new connectors get it from the checklist + one method.
Also fixes the pre-existing Slack
syncCustomEmojitest failure (tool mock was missingscheduleRecurring).Tests
twister 17, gmail 107, google 64, slack 28 (now fully green), outlook-mail 46 (drain-handler, migration, and wiring tests rewritten for the platform drain); ms-teams + outlook composite typecheck (no vitest harness in those packages).
🤖 Generated with Claude Code