refactor(trace-analyst): single source of truth for span-timestamp parsing

[drewstone]

What

Remove a duplicated timestamp parser. store-otlp.ts had local epochOrNull/epochMs that duplicated otlp-span.ts's exported spanEpochMillis/compareSpanTime (identical ISO + epoch-millis dialect handling). Import the shared helpers; drop the local copies.

Why

The two OTLP readers (OtlpFileTraceStore and otlpToRunRecords) now both go through the one canonical timestamp helper, so the dialect handling can't drift between them. Net −13 lines, no new surface.

Verification

Behavior-identical (compareSpanTime(a,b) ≡ epochMs(a)-epochMs(b); spanEpochMillis ≡ epochOrNull): full suite 2461 pass / 0 fail, tsc clean, biome check src clean, timestamp regression tests unchanged.

[tangletools]

✅ Auto-approved PR — 35902acc

Blanket team auto-approval is enabled for this reviewer service.
The full PR reviewer audit still runs separately and will publish findings if it detects issues.

_{tangletools · auto-approval · reason: blanket_auto_approve · 2026-06-19T19:20:29Z}

[tangletools]

🟢 Value Audit — sound

Verdict	sound
Concerns	0 (none)
Heuristic	0.0s
Duplication	0.0s
Interrogation	56.0s (2 bridge agents)
Total	56.0s

💰 Value — sound

Replaces a duplicated timestamp parser (epochOrNull/epochMs) in store-otlp.ts with imports of the identical spanEpochMillis/compareSpanTime from otlp-span.ts — the already-canonical source used by otlpToRunRecords. Net −13 lines, no behavioral change.

• What it does: Removes the local epochOrNull and epochMs helper functions (lines 60–75 in the old file) from store-otlp.ts and replaces their 7 call sites with imports of spanEpochMillis and compareSpanTime from otlp-span.ts. The imported functions are byte-identical in logic: spanEpochMillis handles ISO-8601 and epoch-millis-string dialects identically to epochOrNull; compareSpanTime is `(span
• Goals it achieves: Single source of truth for OTLP span-timestamp parsing. The two OTLP readers (OtlpFileTraceStore and otlpToRunRecords) now both route through the same canonical helpers in otlp-span.ts, eliminating the risk of dialect-handling drift between them. otlpToRunRecords already used the shared imports; this change brings store-otlp.ts into alignment with that established pattern.
• Assessment: This is a mechanical, well-scoped deduplication. Every call site maps one-to-one with identical semantics. The file's own lint-level comment at store-otlp.ts:787-790 already stated that the per-line projection lives in ./otlp-span so both consumers read the same vocabulary — this change fulfills that intent for timestamp parsing too. The timestamp-specific test files exist for both modules (`s
• Better / existing approach: none — this is the right approach. otlp-span.ts is the documented canonical module for shared OTLP vocabulary (src/trace-analyst/otlp-span.ts:7-11 explicitly states "Canonical OTLP-flat-line readers shared by every consumer … One parser, one vocabulary"). otlpToRunRecords (src/trace-analyst/otlp-to-run-records.ts:43,48) already imports compareSpanTime and spanEpochMillis from it. The r
• Model: opencode/deepseek/deepseek-v4-pro
• Bridge attempts: 1

🎯 Usefulness — sound

Removes 13 lines of duplicate timestamp-parsing code in store-otlp.ts by importing the canonical spanEpochMillis/compareSpanTime from otlp-span.ts — the shared helpers already used by otlp-to-run-records.ts, exactly the consolidation the codebase's own design comment (store-otlp.ts:788-790) says it

• Integration: All 5 old call sites in store-otlp.ts (lines 181-182, 530-531, 543, 547-548) correctly rewired to the shared imports. The imported functions are already production-hardened: otlp-to-run-records.ts at lines 43, 48, 285, 302, 304, 340, 341, 411 uses the same compareSpanTime/spanEpochMillis. Both OTLP consumers now share a single parse dialect.
• Fit with existing patterns: Fits the established codebase grain exactly. The module header of otlp-span.ts (lines 7-11) states 'One parser, one vocabulary — a divergence between the analyst's view of a trace and the RunRecord projected from it is a class of bug this consolidation removes.' The old local copies were the anomaly; this change removes them and brings store-otlp.ts into alignment with the already-exported shared
• Real-world viability: Behavior-identical at every call site: compareSpanTime(a,b) ≡ epochMs(a)-epochMs(b); spanEpochMillis ≡ epochOrNull (with the minor improvement that spanEpochMillis accepts string|null|undefined vs the old epochOrNull's string-only signature — slightly more robust against null timestamps). The epoch-millis-string / ISO-8601 dual-dialect handling is preserved identically. No new error paths introduc
• Model: opencode/deepseek/deepseek-v4-pro
• Bridge attempts: 1

No concerns — sound change, no better or existing approach found. ✅

---

What this audit checks

It judges the change on its merits — not whether it was tasked out in an issue. Unticketed, fast-moving work is fine; the question is whether the change is good and whether a better or existing approach should be used instead.

Pass	What it asks
Heuristic	Vague title? Whitespace-only or cruft-bearing diff? (content signals only)
Duplication	Do added function/class names already exist elsewhere in the repo?
Value Audit	What does it do? What goal does it achieve? Is it good? Better architecture or already-exists?
Usefulness Audit	Does it integrate and fit? Will it hold up in real use and actually get used?

Findings are concerns, not blocks — the human reviewer decides what to do with them.

_{value-audit · 20260619T192311Z}

[tangletools]

✅ No Blockers — 35902acc

Readiness 95/100 · Confidence 65/100 · 0 findings (none)

	deepseek	glm	aggregate
Readiness	95	95	95
Confidence	65	65	65
Correctness	95	95	95
Security	95	95	95
Testing	95	95	95
Architecture	95	95	95

Full multi-shot audit completed 1/1 planned shots over 1 changed files. Global verifier still owns final merge decision. | Full multi-shot audit completed 1/1 planned shots over 1 changed files. Global verifier still owns final merge decision.

No findings.

---

_{tangletools · 2026-06-19T19:25:23Z · trace}