3d323bf55fa6b70eb7358665301025441401bbc4
5 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
| 3d323bf55f |
chore(test): switch test database from SQLite to MySQL (WS-6)
Test infrastructure now uses the same MySQL 8.0 engine as local dev
and production. SQLite is no longer used anywhere in the project.
Eliminates the SQLite "rebuild on FK add" quirk that forced session 2.5
to omit a foreign key on form_schemas.default_crowd_type_id (Task 2 of
this session restores it).
Configuration:
- phpunit.xml: DB_CONNECTION=sqlite (:memory:) replaced with mysql
pointing at crewli_test database (127.0.0.1:3306, crewli/secret)
- Makefile: new test-db-create target creates crewli_test in the
bm_mysql Docker container; make test ensures it exists before
running suite
Latent-bug surfacing — fixes that MySQL exposed:
1. form_submissions.idempotency_key was declared `ulid()` (VARCHAR 26)
while FormRequest validates `string|max:30`. SQLite ignored the cap;
MySQL truncated and rejected. Column widened to string(30) to match
validation.
2. FormFieldValidationRuleService / FormFieldConfigService /
FormFieldBindingService::snapshotShapesFor — toJsonShape iterated
collection in DB-default order (insertion-stable on SQLite, undefined
on MySQL). Schema_snapshot bytes drifted across re-emits, breaking
audit-replay. Added `->sortBy('id')` (ULID = insertion-order
semantics, deterministic) on all three.
3. FormSubmissionObserverTest::test_denormalized_indexes_exist queried
sqlite_master directly. Replaced with the cross-engine
information_schema.STATISTICS query (the real production check is
on MySQL anyway).
4. JSON column key order non-determinism: MySQL JSON columns may
round-trip associative-array keys in a different order than they
were inserted. assertSame on JSON-derived associative arrays now
uses assertEquals (structural equality) where the test was previously
over-asserting on key order:
- ConditionalLogicActivityLogPayloadTest
- ConditionalLogicBackfillTest::test_rollback_reconstructs_canonical_json
- FormFieldBindingMigrationTest::test_rollback_reconstructs_json_and_drops_table
- FormFieldOptionServiceAndScopeTest::test_replace_options_emits_activity_log_on_field_only
- FormFieldOptionsActivityLogTest::test_field_updated_payload_contains_options_diff_when_options_change
- FormFieldOptionsBackfillTest::test_forward_migration_backfills_rows_strips_translations_and_rewrites_snapshot
- FormFieldOptionsSnapshotAndStrictRequestTest::test_submission_snapshot_embeds_rich_shape_options
5. Backfill / migration tests (4 classes, 21 tests) ran migrate:rollback
then migrate inside RefreshDatabase's wrapping transaction. MySQL
DDL implicit-commits the surrounding transaction, leaving Laravel
unable to ROLLBACK TO SAVEPOINT at end-of-test (1305 SAVEPOINT
does not exist). Replaced RefreshDatabase with a per-test
migrate:fresh in setUp + RefreshDatabaseState::\$migrated = false to
force the next RefreshDatabase test to re-migrate cleanly:
- FormFieldBindingMigrationTest
- ConditionalLogicBackfillTest
- FormFieldOptionsBackfillTest
- FormFieldValidationRuleBackfillTest
All 1386 tests now pass on MySQL. Larastan baseline unchanged.
Refs: WS-6 session 2.5 deviation #1 cleanup, RFC-WS-6.md v1.1
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
|||
| e7c9482474 |
refactor(form-field): drop form_fields.options + form_field_library.options
Final WS-5d cleanup. The JSON columns that have been unread since
commit 3 are now physically dropped on both source tables. Their
canonical rich-shape lives in form_field_options, accessed
exclusively through the morphMany relation.
Defensive sweep: any lingering translations.{locale}.options key in
either source table's translations bag is stripped. Commit 2's
backfill should already have done so exhaustively; this is
belt-and-braces.
Rollback re-creates the columns as nullable JSON but leaves them
empty. Pair with commit 2's rollback to restore the pre-WS-5d data
shape on every owner row.
The commit-3 getOptionsAttribute accessor-bridge on FormField +
FormFieldLibrary is removed — Eloquent's getAttribute() resolution
now naturally falls through to the morphMany relation since there's
no underlying column to shadow it. New regression test
FormFieldOptionsAccessTest asserts $field->options resolves to an
Eloquent Collection of FormFieldOption instances and lazy-loads in
exactly 2 queries (1 parent + 1 lazy-load options) on a fresh fetch
without with() preload. Same trio for FormFieldLibrary.
Migration step-count tests in WS-5a/b/c bumped by 1 to account for
the new drop_form_field_options_json_columns migration on the
rollback stack.
Documentation:
- SCHEMA.md v2.6: form_field_options table documented; options row
removed from form_fields and form_field_library; morphMany
relations updated; cross-references to ARCH-FORM-BUILDER §17.6
and addendum §Q3 WS-5d Uitvoering added on both source-table
docblocks.
- ARCH-FORM-BUILDER.md v1.8: new §17.6 "Field options (relational)"
mirrors the §17.4 / §17.5 relational-sibling structure with
sub-sections 17.6.1 rationale, 17.6.2 table + catalogue, 17.6.3
service / scope / cascade / activity log, 17.6.4 snapshot
embedding, 17.6.5 external API contract. Existing Webhooks
section renumbered from §17.6 to §17.7.
- ARCH-CONSOLIDATION-ADDENDUM-2026-04-24.md: "Uitvoering — WS-5d
(2026-04-27)" section added. Eight paragraphs covering the
snapshot atomic rewrite, strict-fail backfill dispatch, dual
activity-log emit, four-sibling base-class extraction warrant,
commit 0 dead-code precondition, the temporary getOptionsAttribute
accessor-bridge pattern (with reusability note for future
JSON→relational refactors), the dev-seeder vergoedingstype RADIO
normalisation (drift correction explicitly distinguished from the
parallel apps/app RegistrationFieldTemplate description domain),
and the WS-5 family completion note.
- BACKLOG.md: FORM-BUILDER-LIBRARY-AUDIT-LOG entry extended to four
services (adds library.options_replaced); new
FORM-BUILDER-MORPH-SCOPE-BASE-CLASS entry added as the WS-5d
follow-up now that all four concrete morph-scope siblings exist.
Tests: 1193 → 1208 green (+15 across commits 3+4+5; this commit alone:
+2 from the regression test).
This completes the WS-5 family.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
|||
| bb9242fd6e |
refactor(form-field): resources + snapshot + validator read form_field_options
Atomic reader switch. All call paths that previously read
form_fields.options / form_field_library.options from the JSON column
now read through FormFieldOptionService::toJsonShape() via the
morphMany relation:
- FormFieldResource + FormFieldLibraryResource +
PublicFormSchemaResource emit the rich-shape array
- FilterRegistryController emits rich shape uniformly (no flat-array
carve-out for filter-UI compatibility — preflight scan confirmed
zero portal/app consumers, S5 territory)
- FormFieldRuleBuilder plucks values from the relation for in:options
rule construction
- FormSubmissionService::buildSnapshot writes rich-shape options into
snapshots and strips translations.{locale}.options from each field's
translations bag (defensive — commit 2 backfill already did the
bulk strip)
- Four FormFieldRequest variants accept array-of-spec-objects,
validate shape in after() via FormFieldOptionService::assertSpecsValid,
and hand off to FormFieldOptionService::replaceOptions for writes
- FormFieldService::create + update extract option specs from the
request data and route through the service after the FormField row
is persisted
FormField and FormFieldLibrary $casts no longer include 'options'; the
JSON column is no longer cast. Options removed from $fillable on both
models so ::create() / ::fill() / mass assignment can no longer touch
the legacy column. Both models gain a getOptionsAttribute() accessor
that resolves $model->options to the eager-loaded morphMany collection
— required because Eloquent's getAttribute() prefers a real DB column
over a relation method, and the JSON column lives on the table until
WS-5d commit 5 drops it.
Activity log — dual emit per §6.7 / §17.4.2 / §17.6.3:
- field.updated carries old.options / new.options diff via
toJsonShape() reconstruction, byte-equal JSON compare to avoid
cosmetic false positives. Field updates that don't touch options
omit the key entirely
- field.options_replaced emits inside replaceOptions() on FormField
subject only; library subject writes silent (mirrors the WS-5b /
WS-5c convention)
JSON columns (form_fields.options, form_field_library.options) remain
present but unread — column drops land atomically in commit 5.
Two pre-existing test fixtures that seeded options via the JSON column
(FormFieldApiTest + PublicFormValidationTest) migrated to the
spec-array path: FormField::factory()->withOptions([...]) where the
options live on the field, or explicit spec-array request bodies for
HTTP tests.
Tests: 1193 → 1206 green (+13 tests / +28 assertions).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
|||
| 15e4e49d8c |
feat(form-field): backfill form_fields.options to form_field_options
Atomic data migration. Every options datum in the database — in
form_fields and form_field_library, their translations bags, and the
form_submissions.schema_snapshot + form_templates.schema_snapshot JSON
blobs — is converted to the new relational rich-shape representation.
Strict dispatch per §17.4.4 / §8.7 convention:
- Fail on field_type ∉ {RADIO, SELECT, MULTISELECT, CHECKBOX_LIST}
carrying non-null options (post-WS-5b TAG_PICKER seed-bug indicator)
- Fail on non-flat-string-array options shape
- Fail on translations.{locale}.options[] length mismatch
- Fail on non-string / >255-char translated labels
- Fail on any residual translations.{locale}.options key after
step C migration
Snapshot rewrite in-place: both form_submissions.schema_snapshot and
form_templates.schema_snapshot walk fields[*] and rewrite options to
the new rich-shape, strip per-locale options[] from the parallel
translations bag. Zero-compromise directive — no reader tolerance for
pre-WS-5d shape in commit 3 onwards.
Rollback reconstructs JSON column shapes plus translations bags.
Forward+back pair safe as a unit; partial rollback unsupported.
FormFieldService::insertFromLibrary switches from JSON-copy to
FormFieldOptionService::copyOptions row-clone per addendum Q3 row-copy
mandate. The field's own translations bag no longer carries
{locale}.options keys — those live on option rows now.
Seeders and factories switch to service-level option creation:
- FormBuilderDevSeeder.canonicalFields keeps flat-string options as
its data shape; FormField::create no longer receives an options
key, the post-create FormFieldOptionService::replaceOptions call
inserts the rich rows. The same applies to
seedEventRegistrationShowcaseSchema. The vergoedingstype field's
legacy {label, description} object shape (a pre-WS-5d seed-bug
that the strict backfill would reject) is normalised to flat
strings; the descriptions are dropped.
- seedSystemTemplates embeds rich-shape options in the template
snapshot — no flat-array snapshot data remains in newly-seeded
rows.
- FormFieldFactory + FormFieldLibraryFactory drop the options
default; new ::withOptions() helper accepts either flat strings
(each becomes value+label) or full spec arrays and routes through
the service.
JSON columns (form_fields.options, form_field_library.options) remain
present and writable via fillable; column-drop lands in commit 5.
Reads from the JSON column still exist in resources, snapshot writer,
FormRequests, FormValueService, and FilterRegistryController — commit
3 switches those all atomically.
Migration step-count tests in WS-5a/b/c bumped by 1 to account for
the new backfill_form_field_options migration on the migration stack.
Tests: 1182 → 1193 green (+11 tests / +56 assertions).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
|||
| 11588623c5 |
feat(form-field): add form_field_options table, service, scope, cascade
Fourth and final WS-5 sibling. Polymorphic morph-owned table for the
RADIO / SELECT / MULTISELECT / CHECKBOX_LIST option rows, shared
between form_fields and form_field_library via the owner_type
discriminator. Matches the WS-5a (bindings) / WS-5b (validation_rules
+ configs) pattern one-for-one: dedicated service as single writer,
UNION-over-two-owner-chains scope, shared cascade observer.
Row shape:
- value canonical storage value (string ≤255, UNIQUE per owner)
- label default-locale display label (string ≤255)
- sort_order int unsigned
- translations JSON { "<locale>": "<translated label>" }
The UNIQUE(owner_type, owner_id, value) index ffo_owner_value_unique
is the seed-bug guard — duplicate values per field have no semantic
meaning and must fail at both the service layer (assertSpecsValid)
and the DB level.
Activity log: field.options_replaced emits on FormField subject only,
per the §6.7 WS-5a / §17.4.2 WS-5b convention that library-level
changes are silent in activity log.
No production reads yet. The form_fields.options and
form_field_library.options JSON columns remain the active source of
truth until the commit-3 reader switch — accessing $field->options
still resolves through the JSON cast in commit 1, so model tests
exercise the new morphMany via $field->options() (explicit relation
call). Both FormField and FormFieldLibrary now carry an `options`
morphMany alongside `bindings`, `validation_rules`, and `configs`.
Cascade: FormFieldChildTablesCascadeObserver gains form_field_options
as the fourth child cleaned on owner delete (both FormField soft/
force-delete and FormFieldLibrary delete).
Migration step-count tests in WS-5a/b/c bumped by 1 to account for
the new create_form_field_options_table on the migration stack.
Base scope-class extraction across the four siblings — deliberately
deferred to a follow-up work package per addendum §17.4.3 / §17.5.3.
Now that all four concrete implementations exist, the "what actually
varies" question can be answered empirically.
Tests: 1158 → 1182 green (+24 tests / +42 assertions).
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
|
