Repository navigation
[PR] perf: scan native compaction SSE frames in linear time - #6579
Merged
ranxianglei merged 1 commit intoNov 14, 2026
Merged
Conversation
This was referenced Nov 14, 2026
Owner
|
Superseded by #6677 (merged as Landed into master via the real merge commit The fix for #6578 is now in master. Closing this PR as superseded — thanks for the careful, well-tested fix. |
This was referenced Nov 15, 2026
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 free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes #6578
Problem
The native compaction observer appends each network chunk to an accumulating string, then normalizes and scans the whole unfinished frame again. A response.completed event can include the complete compacted response, so a long data field repeatedly copies and searches its growing prefix on the relay's event loop.
Changes
Scan each newly decoded chunk once. Keep incomplete-line fragments until a line terminator arrives, join each line once, and assemble the event once at its blank line. A single pending-CRLF flag preserves line-ending behavior across chunk boundaries. Existing terminal classification remains the authority for the first complete verdict.
Validation
node --import tsx scripts/bench-terminal-sse.ts 8 32. The benchmark has no timing assertion in CI.Base: e1f31ad. Checked previous Afloat16 PRs and fresh all-state terminal/SSE searches. Existing dispatch-completeness and JSON-status fixes retain their coverage. Hosted CI and real-model E2E remain separate merge gates.