Root cause: child model root_validators (Vision ID anti-leakage guard) strip
integer IDs before they can be serialized into the INSERT dict, causing MariaDB
to reject the INSERT with 'Field does not have a default value' (1364).
Fix: re-inject resolved_parent_id into data_to_insert after validated_obj.dict()
in post_child_obj(). This is safe — the integer was already verified against the
DB before model validation.
Affected (were all broken since ~2026-01-27):
- journal/{id}/journal_entry/
- event/{id}/event_session/
- event/{id}/event_person/
- event/{id}/event_registration/
- event/{id}/event_presenter/
- event/{id}/event_presentation/
- event/{id}/event_location/
- event/{id}/event_track/
- event/{id}/event_device/
- event/{id}/event_abstract/
- event/{id}/event_badge/ (different symptom: NULL FK)
Tests: add nested create lifecycle regression tests to test_e2e_v3_demo_parity.py
- POST + Vision check + DELETE for journal/journal_entry and event/event_session
- All 9 checks passing (7s)
Docs: update tests/README.md with accurate demo_parity description and
a 'When to Run Tests' matrix to prevent future gaps in coverage.
Address critical data visibility issues for Event Files and enhance frontend documentation.
This commit resolves the persistent problem where top-level hosted file convenience fields
(e.g., , , ) were
returning as in V3 Event File API responses, even when .
Key changes include:
- Refactored Pydantic model:
- Removed redundant definitions from top-level hosted file convenience fields,
allowing direct mapping from SQL view columns.
- Simplified to focus solely on conditionally loading the nested
object, as top-level fields are now populated directly by Pydantic
from the view.
- Added comprehensive comments to clarify data flow, Pydantic's behavior, and the
expected origin of these convenience fields from SQL views.
- Updated :
- Introduced a new section detailing how to retrieve Event File data, including the
use of to get both top-level convenience fields and a nested
object.
- Clarified all ID references as random string IDs.
- Renumbered the troubleshooting section.
- Copied updated guide to .
- Continued ID Vision compliance audit, ensuring consistent handling of random string IDs
across various core and event models (Account, Address, Contact, DataStore, Event Badge Template).
- Consolidated ID Vision E2E tests and updated related documentation.
- Minor updates to and
to support Event File data retrieval with .
1. Added 'account_id' and 'account_id_random' to searchable_fields for event_exhibit_tracking.
2. Updated tests/README.md with descriptions for the latest E2E test scripts (Demo Parity, Event Actions, Zoom).
- Combined 10+ one-off tests into 4 primary functional suites (Search, Auth, Lifecycle, Vision).
- Archived original scripts to tests/archive/.
- Updated README with the new standardized inventory.
- Applied clean output formatting across the new suite.
- Moved legacy/redundant tests to tests/archive/.
- Relocated root-level debug scripts to tests/integration/.
- Updated tests/README.md with final organized inventory.
- Cleaned up root directory from one-off reproduction scripts.
- Introduced WS_Message_V3 standardized Pydantic model and WS_Manager_V3.
- Implemented /v3/ws/ endpoint with granular Redis routing to solve "noisy neighbor" scaling issues.
- Added presence tracking using Redis Sets for group coordination.
- Comprehensive test suite added (unit and integration) covering models, manager, and routing logic.
- Documentation: Created V3 Frontend WebSocket Guide and Project design spec.
- Updated main Frontend API guide and tests README with new standards.
- Patched request_jwt to strip privileged IDs when signing with public keys
- Updated AccountContext and V3 dependencies to preserve JWT payloads for guests
- Whitelisted Archive, Post, Event, and other core objects for public read access
- Added 'default_qry_str' to Event searchable fields
- Added test_e2e_jwt_guest_auth.py for security verification