fix(persistence,rest): _lastUpdated on Postgres; server-populated meta in response bodies - #876
Merged
Merged
Conversation
…nse bodies Two findings from the #448 Postgres leg. _lastUpdated searches returned an opaque 500 on the Postgres backend at every precision and prefix: build_last_updated_condition bound the raw query string as a text parameter against the TIMESTAMPTZ last_updated column, which tokio-postgres refuses to serialize, and the path had none of the precision-range semantics the date parameters got in #463. It now binds real timestamps and mirrors build_date_condition: eq at day precision means [day, day+1), gt excludes the named period, and a full-precision instant falls back to scalar comparison. Response bodies carried no server-populated meta on any read path — versionId and lastUpdated lived only in the ETag and Last-Modified headers, so search entries, reads, vreads, history bundles, and batch/transaction responses returned resources with no server metadata, and the Resources workspace's UPDATED column rendered blank on every backend. StoredResource::content_with_meta now merges the row's version and timestamp into the returned body (client-supplied meta members like profile survive), and every handler and bundle builder that echoed stored content uses it. Verified live against postgres:16 with the #448 dataset: the full battery passes 22/22, and a Patient read returns meta.profile + versionId + lastUpdated together. Closes #871 Closes #873
9 tasks
Resolve conflict in core/search.rs: main split SearchResult::to_bundle into a shared bundle_shell plus a copying to_bundle and a consuming into_bundle. Apply the #873 server-populated meta merge to both paths, adding a consuming StoredResource::into_content_with_meta so into_bundle keeps its clone-free behaviour. Claude-Session: https://claude.ai/code/session_01334x1QZSmUsznAtEVcnVy7
This was referenced Sep 2, 2026
test_prefix_operators asserted the scalar operators the pre-#871 code emitted — the exact semantics the fix replaces. Day-precision prefixes now assert their [day, day+1) range shapes and bound counts, and a full-precision instant pins the scalar fallback.
Contributor
Author
|
CI fix pushed: |
…iosSoftware/hfs into fix/871-873-read-correctness
… manifest `postgres_bulk_submit_import_directives_round_trip` called `claim_next_manifest` and asserted the claimed view carried the directives it had just set. The claim queue is cross-tenant and ordered by `added_at`, the test binary shares one container database, and the batch test that #880 added leaves its manifest as `processing` with no lease — which the claim query treats as an orphan to reclaim. Whenever that manifest was added first, the directives test claimed it instead of its own and failed with `left: []`; main's coverage job has been red on most runs since. The test now claims through `claim_specific_manifest`, the submit-side twin of the export tests' `claim_specific`: loop until the target manifest comes back, hold any foreign lease picked up along the way so it cannot be re-claimed, then release those back to the queue. The batch test's comment no longer presents its `process_entries` call as a defence against concurrent claims. No product code changes. Tests: the two submit tests pass five consecutive runs together; the full postgres_tests binary passes (156).
Contributor
Author
|
CI here was failing on |
Codecov Report❌ Patch coverage is 📢 Thoughts on this report? Let us know! |
smunini
approved these changes
Sep 3, 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
3 participants
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.
Closes #871
Closes #873
Both found by the #448 Postgres sanity leg; base is
main(independent of the bulk-progress stack).#871 —
_lastUpdated500s on Postgresbuild_last_updated_conditionbound the raw query string as a text parameter against theTIMESTAMPTZlast_updatedcolumn — tokio-postgres refuses to serialize it ("error serializing parameter 2"), so every precision and prefix returned an opaque 500. The path also predated the precision-range semantics dates got in #463.Now it binds real timestamps and mirrors
build_date_conditionexactly:eqat day precision means[day, day+1),neits complement,gt/saexclude the named period,lt/le/geuse the correct boundary, and a full-precision instant falls back to scalar comparison. Unit tests pin the SQL shape, the bound param types, and the period-end arithmetic.#873 — response bodies carried no server-populated
metaversionId/lastUpdatedlived only in theETag/Last-Modifiedheaders: search entries, reads, vreads, history bundles, and batch/transaction responses all returned resources with no server metadata — R4 requires it at least on vread, Bundles have no per-resource headers, and the Resources workspace's UPDATED column rendered blank on every backend.Fix is read-side and central:
StoredResource::content_with_meta()merges the row's version and timestamp into the returned body — client-suppliedmetamembers (profile,tag,security) survive,versionId/lastUpdatedare overwritten because the row is authoritative — and every echo point uses it: read, vread, history entries, searchto_bundle(match + include), create/update/patch responses, batch/transactionBundleEntryResult, and subscription event payloads. Read-side means pre-existing rows are fixed immediately, no migration.An integration test walks the whole lifecycle: create (v1 meta in the response, profile preserved) → update → read (v2) → vread (v1 keeps its version) → history (per-version meta) → search (entries carry meta).
Verified
meta.profile + versionId + lastUpdatedtogether.