Commit Graph
4 Commits
Author SHA1 Message Date
DongHyeonka 348420618d fix: stop claiming a 1x1 size for an image whose dimensions are unknown
An uploaded figure never appeared, and no request for it was ever made. The
resolver substituted 1x1 when an asset carried no dimensions, and a 1x1 box
with `loading="lazy"` never enters the viewport — so the browser had no reason
to fetch it. The image was not failing to load; it was never asked for.

Unknown dimensions now say so. The figure omits the attributes and loads
eagerly, letting the browser size the image from the file, and reserves layout
space only when the size is actually known. Guessing a number to fill an
attribute is what turned a missing measurement into a missing picture.
2026-08-21 17:57:10 +09:00
DongHyeonkaandClaude Opus 5 073fda87eb fix: collapse the evidence-key gate and resolver into one decision
Three fix rounds each rebuilt the gate as a separate expression that merely
agreed with the resolver on the inputs that round's tests used. Different
expressions cannot agree in general, so the defect class stayed open while
each reported instance closed.

`findResolvableAsset(assets, key)` is now the single place that decides which
Asset an evidence key resolves to. Every gate is
`Boolean(findResolvableAsset(...)) || legacyKey(key)` via one shared
composition, and every resolver returns what it returns:

- validate-working-copy: the key gate and the decorative lookup (a last-wins
  Map against the resolver's first-wins find, so alt could be judged against a
  different Asset than the one rendered)
- adapters/mock/project-public-render-model: gate and resolver
- instant-preview: gate and descriptor resolver
- createAssetCatalogResolver: the pixels, a fourth expression nobody had
  listed -- one Asset's caption could sit over another Asset's image

Duplicate assetKeys are a contract violation but reachable through a paged
list, so the choice is total and order-independent: newest updatedAt wins,
tie-broken by id.

InstantPreview's gate is no longer looser than the others. The un-loaded-asset
case it was loosened for blanks either way; all the looseness bought was
catalog-only keys rendering an empty gap with no message while validation said
EVIDENCE_UNSUPPORTED. The test that pinned that divergence now asserts the
consistent behaviour, and the false comment claiming a fix that did not exist
is gone.

idempotent() now maps a deterministic content failure to VALIDATION_STALE/409
instead of offering a retry that fails identically, and no longer caches
uncharacterized internal failures -- reporting one as retryable while freezing
it in the ledger meant the retry could never re-run.

Adds tests/features/tech-log/evidence-key-agreement.test.ts: 144 adversarial
(asset list, key) combinations asserting the agreement itself rather than
examples. It reported 53 disagreements against the previous code.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-18 06:55:06 +09:00
DongHyeonkaandClaude Opus 5 e2c1d076f4 fix: overlay legacy evidence labels instead of shadowing the whole descriptor
Review finding I1: the static-key-first resolution order protected the
entire legacy fixture (image and labels), when only the hand-authored labels
ever needed protecting -- ResolvedAsset carries no label fields, so nothing
else could recover them, but a real descriptor's publicPath/width/height were
never actually at risk of mismatching. Narrowed resolveWith so a resolved
descriptor's data fields always win; the legacy registry only overlays
triggerLabel/dialogLabel for keys it recognizes, and only supplies the full
descriptor when nothing else resolves the key at all. A future backend asset
colliding with the legacy key now degrades to a wrong caption, never a wrong
image.

Review finding I2: renamed and rewrote a test whose title claimed a
READY-vs-QUARANTINED same-key guarantee its body never constructed. It now
puts both a QUARANTINED and a READY entry under one assetKey and asserts the
READY one wins.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-18 04:03:43 +09:00
DongHyeonkaandClaude Opus 5 9d91001a31 feat: resolve evidence figures from backend asset descriptors
Public Preview and Publication Snapshot now resolve evidence figures from the
ResolvedAsset descriptor the server already attaches to each EVIDENCE_FIGURE
block, instead of a local literal that only knew one hardcoded key and threw
on anything else. Instant Preview gains an `assets` prop (defaults to `[]`,
unwired until the Asset Picker task) and stops throwing when a key has no
catalog match yet.

Builds on Task 1's existing seam (resolveCaseEvidenceAssets /
ResolveEvidenceAsset) via a new asset-resolvers.ts rather than a parallel
path. Kept out of adapters/ (presentation may not import it) by carrying a
small local copy of the one legacy fixture entry, checked before any
descriptor match so the pre-existing key keeps rendering byte-identically
across surfaces during the migration window. Non-READY assets never resolve;
unresolvable keys return a placeholder instead of throwing.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-08-18 03:54:30 +09:00