Skip to content

feat(planner): accept optional measured_at on atomic-cost entries - #796

Merged
zzylol merged 1 commit into
mainfrom
atomic-costs-measured-at
Oct 6, 2026
Merged

zzylol merged 1 commit into
mainfrom
atomic-costs-measured-at

Conversation

@zzylol

@zzylol zzylol commented Oct 6, 2026

Copy link
Copy Markdown
Contributor

ProjectASAP/sketch-bench#155 (part of sketch-bench#147) adds a measured_at block to every row approxbench atomic-costs writes: items, keys and value range per instance, the merge operand size, and the value distribution.

AtomicCostEntry here has deny_unknown_fields, so a table regenerated after that PR would fail to load. This adds measured_at: Option<serde_json::Value>, #[serde(default)] and omitted when absent. It is carried opaquely, like a synthetic workload description, since the planner doesn't use it yet.

Merge before regenerating the atomic-cost table with a sketch-bench that includes #155.

Test plan

  • new test: an entry with measured_at loads, one without still loads, any other unknown field is still refused
  • cargo test -p asap_planner --lib (136 passed), cargo clippy -p asap_planner --lib --tests -- -D warnings

🤖 Generated with Claude Code

sketch-bench#155 (for sketch-bench#147) adds a measured_at block to every
atomic-cost row. AtomicCostEntry denies unknown fields, so a regenerated
table would fail to load. Carry it as an optional opaque Value.

Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
@milindsrivastava1997 milindsrivastava1997 changed the title planner: accept optional measured_at on atomic-cost entries feat(planner): accept optional measured_at on atomic-cost entries Oct 6, 2026
@zzylol
zzylol merged commit 35a1ee1 into main Oct 6, 2026
8 of 9 checks passed
@zzylol
zzylol deleted the atomic-costs-measured-at branch October 6, 2026 19:27
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants