fix(form-builder): canonicalize JSON for byte-stable storage (WS-6)

MySQL 8.0 JSON columns may reorder associative-array keys on
round-trip. For audit-immutable values (schema snapshots, webhook
payloads, activity log diffs), this is corrupting: re-emits produce
different byte sequences for the same logical content.

Introduced JsonCanonicalizer (recursive ksort on associative arrays;
numeric-indexed lists preserve order) and applied at every writer
site that produces byte-stable JSON:

- FormSubmissionService: canonicalize the schema_snapshot array
  before storage (audit-immutable per ARCH §4.3, RFC-WS-6 v1.1).
- FormField::logFieldChange / FormSchema::logSchemaChange: canonicalize
  activity-log properties before withProperties() so old/new diffs
  read back byte-stable.
- BindingActivityLogger: canonicalize both the pass-level and
  per-binding activity properties.
- FormWebhookDispatcher: canonicalize payload_snapshot before
  storage (delivery-time HMAC re-encodes the same canonical bytes).
- DeliverFormWebhookJob: switched json_encode to
  JsonCanonicalizer::encode for the HMAC-signed body, so the
  signature is byte-stable across re-deliveries and reproducible by
  receivers from the same logical payload.

Sites NOT canonicalized (deliberate):
- form_schemas.settings — opaque UI config; key order has no
  semantic meaning, no byte-stability requirement.
- form_schemas.translations / form_fields.translations — read by
  display layer; key order doesn't matter.
- form_templates.schema_snapshot — user-supplied input via store/
  update; user is the source of truth, not audit-immutable in the
  same way as form_submissions.schema_snapshot.

Reverted the 7 assertEquals workarounds from session 2.6:
- 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

Each now uses assertSame on JsonCanonicalizer::encode of both sides —
byte-stable comparison meaningful regardless of MySQL JSON storage
behavior.

New regression test SchemaSnapshotByteStableAcrossReemitsTest
exercises the contract end-to-end: complex schema with bindings,
validation rules, options, conditional logic, submitted; reads
schema_snapshot via three roads (Eloquent cast, fresh model, raw
bytes) and asserts the canonical encode is identical.

ARCH-FORM-BUILDER.md §4.6.1 gets a "Byte-stability" sub-section
explaining what's canonicalized and why.

Test count: 1388 → 1400 (+11 JsonCanonicalizer unit, +1 snapshot
regression). Larastan clean. Rector dry-run unchanged at 355.

Refs: WS-6 session 2.6 deviation #4 cleanup, RFC-WS-6 v1.1

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-04-28 13:51:38 +02:00
parent 0afbd36bf7
commit a791a276fa
17 changed files with 488 additions and 82 deletions

View File

@@ -5,6 +5,7 @@ declare(strict_types=1);
namespace App\FormBuilder\Bindings;
use App\Models\FormBuilder\FormSubmission;
use App\Support\Json\JsonCanonicalizer;
use Spatie\Activitylog\Models\Activity;
/**
@@ -23,9 +24,12 @@ final class BindingActivityLogger
{
public function logPass(FormSubmission $submission, BindingPassResult $result): void
{
// RFC-WS-6 session 2.7 — canonicalize properties before they land
// in activity_log.properties (MySQL JSON column round-trip would
// otherwise reorder keys and break diff/regression assertions).
$passActivity = activity()
->performedOn($submission)
->withProperties([
->withProperties(JsonCanonicalizer::canonicalize([
'binding_count' => count($result->applications),
'succeeded' => $result->successCount(),
'failed' => $result->failureCount(),
@@ -33,7 +37,7 @@ final class BindingActivityLogger
'person_provisioned' => $result->provisionedSubjectType === 'person',
'subject_type' => $result->provisionedSubjectType,
'subject_id' => $result->provisionedSubjectId,
])
]))
->log('form_submission.bindings_pass_completed');
$parentActivityId = $passActivity instanceof Activity ? (string) $passActivity->id : null;
@@ -56,7 +60,7 @@ final class BindingActivityLogger
activity()
->performedOn($submission)
->withProperties($properties)
->withProperties(JsonCanonicalizer::canonicalize($properties))
->log('form_submission.binding_applied');
}
}

View File

@@ -6,6 +6,7 @@ namespace App\Jobs\FormBuilder;
use App\Enums\FormBuilder\FormWebhookDeliveryStatus;
use App\Models\FormBuilder\FormWebhookDelivery;
use App\Support\Json\JsonCanonicalizer;
use Illuminate\Bus\Queueable;
use Illuminate\Contracts\Queue\ShouldQueue;
use Illuminate\Foundation\Bus\Dispatchable;
@@ -61,8 +62,13 @@ final class DeliverFormWebhookJob implements ShouldQueue
return;
}
// RFC-WS-6 session 2.7 — canonical JSON for HMAC signing.
// payload_snapshot was read from a MySQL JSON column whose key
// order may not match what we wrote. Canonicalize so the
// signature is byte-stable across re-deliveries and matches what
// a verifying receiver computes from the same logical payload.
$payload = (array) ($delivery->payload_snapshot ?? []);
$body = json_encode($payload, JSON_THROW_ON_ERROR);
$body = JsonCanonicalizer::encode($payload);
$headers = ['Content-Type' => 'application/json'];
if (! empty($webhook->secret)) {
@@ -173,7 +179,7 @@ final class DeliverFormWebhookJob implements ShouldQueue
}
$maskLong = -1 << (32 - (int) $mask);
return (($ipLong & $maskLong) === ($subnetLong & $maskLong));
return ($ipLong & $maskLong) === ($subnetLong & $maskLong);
}
private function isRetriable(int $status): bool

View File

@@ -7,6 +7,7 @@ namespace App\Models\FormBuilder;
use App\Enums\FormBuilder\FormFieldDisplayWidth;
use App\Enums\FormBuilder\FormValueStorageHint;
use App\Models\Scopes\OrganisationScope;
use App\Support\Json\JsonCanonicalizer;
use Illuminate\Database\Eloquent\Concerns\HasUlids;
use Illuminate\Database\Eloquent\Factories\HasFactory;
use Illuminate\Database\Eloquent\Model;
@@ -32,7 +33,7 @@ final class FormField extends Model
protected static function booted(): void
{
static::addGlobalScope(new OrganisationScope());
self::addGlobalScope(new OrganisationScope);
}
/** @return array{via: class-string, fk: string} */
@@ -156,9 +157,13 @@ final class FormField extends Model
*/
public function logFieldChange(string $event, array $properties = []): void
{
// RFC-WS-6 session 2.7: properties land in activity_log.properties
// (MySQL JSON column). Canonicalize so diff/regression assertions
// and downstream consumers see byte-stable structure regardless of
// MySQL key-order normalization on round-trip.
activity()
->performedOn($this)
->withProperties($properties)
->withProperties(JsonCanonicalizer::canonicalize($properties))
->log($event);
}
}

View File

@@ -11,6 +11,7 @@ use App\Models\CrowdType;
use App\Models\Organisation;
use App\Models\Scopes\OrganisationScope;
use App\Models\User;
use App\Support\Json\JsonCanonicalizer;
use Illuminate\Database\Eloquent\Concerns\HasUlids;
use Illuminate\Database\Eloquent\Factories\HasFactory;
use Illuminate\Database\Eloquent\Model;
@@ -33,7 +34,7 @@ final class FormSchema extends Model
protected static function booted(): void
{
static::addGlobalScope(new OrganisationScope());
self::addGlobalScope(new OrganisationScope);
}
protected $fillable = [
@@ -152,9 +153,13 @@ final class FormSchema extends Model
*/
public function logSchemaChange(string $event, array $properties = []): void
{
// RFC-WS-6 session 2.7: properties land in activity_log.properties
// (MySQL JSON column). Canonicalize so diff/regression assertions
// and downstream consumers see byte-stable structure regardless of
// MySQL key-order normalization on round-trip.
activity()
->performedOn($this)
->withProperties($properties)
->withProperties(JsonCanonicalizer::canonicalize($properties))
->log($event);
}
}

View File

@@ -19,9 +19,9 @@ use App\Models\FormBuilder\FormSubmission;
use App\Models\FormBuilder\FormSubmissionDelegation;
use App\Models\FormBuilder\FormValue;
use App\Models\User;
use App\Support\Json\JsonCanonicalizer;
use Illuminate\Database\Eloquent\Model;
use Illuminate\Support\Facades\DB;
use Illuminate\Support\Str;
/**
* Submission lifecycle: draft submitted reviewed per ARCH §4.3, §15.
@@ -103,7 +103,7 @@ final class FormSubmissionService
{
$this->assertWritable($submission);
$result = DB::transaction(function () use ($submission, $actor): FormSubmission {
$result = DB::transaction(function () use ($submission): FormSubmission {
$schema = $submission->schema;
$submission->status = FormSubmissionStatus::SUBMITTED->value;
@@ -111,7 +111,12 @@ final class FormSubmissionService
$submission->schema_version_at_submit = $schema->version;
if ($schema->snapshot_mode !== FormSchemaSnapshotMode::NEVER) {
$submission->schema_snapshot = $this->buildSnapshot($schema);
// RFC-WS-6 session 2.7: schema_snapshot is audit-immutable;
// canonicalize before storage so MySQL JSON-column round-trip
// can never corrupt audit-replay diffs or webhook signing.
$submission->schema_snapshot = JsonCanonicalizer::canonicalize(
$this->buildSnapshot($schema),
);
}
if ($submission->opened_at !== null) {
@@ -272,7 +277,6 @@ final class FormSubmissionService
* any residual options key defensively (commit 2 backfill should
* already have done so on existing rows).
*
* @param mixed $translations
* @return array<string, mixed>|null
*/
private function stripOptionsFromTranslations(mixed $translations): ?array

View File

@@ -9,6 +9,7 @@ use App\Jobs\FormBuilder\DeliverFormWebhookJob;
use App\Models\FormBuilder\FormSchemaWebhook;
use App\Models\FormBuilder\FormSubmission;
use App\Models\FormBuilder\FormWebhookDelivery;
use App\Support\Json\JsonCanonicalizer;
/**
* Finds active webhooks for a submission's schema + trigger and queues a
@@ -36,7 +37,12 @@ final class FormWebhookDispatcher
'trigger_event' => $triggerEvent,
'status' => FormWebhookDeliveryStatus::PENDING->value,
'attempts' => 0,
'payload_snapshot' => $this->buildPayload($submission, $triggerEvent),
// RFC-WS-6 session 2.7 — canonicalize before storage; the
// delivery job HMAC-signs the same canonical bytes after
// re-encode, so signature is reproducible.
'payload_snapshot' => JsonCanonicalizer::canonicalize(
$this->buildPayload($submission, $triggerEvent),
),
]);
DeliverFormWebhookJob::dispatch($delivery->id)->onConnection('webhooks')->onQueue('webhooks');

View File

@@ -0,0 +1,75 @@
<?php
declare(strict_types=1);
namespace App\Support\Json;
/**
* Canonical JSON encoding for byte-stable storage.
*
* MySQL 8.0 JSON columns may reorder associative-array keys on
* round-trip. For values that need byte-stability (schema snapshots,
* webhook payloads signed via HMAC, audit-replay diffs), canonicalize
* the structure before encode so re-emits produce identical bytes.
*
* Strategy:
* - Associative arrays: recursively ksort
* - Numeric-indexed lists (`array_is_list()`): preserve order
* (semantically ordered)
* - Scalars and non-arrays: passthrough
*
* Numeric vs associative detection follows array_is_list() PHP 8.1+
* convention. Mixed-key arrays are treated as associative (rare and
* indicative of a data issue worth surfacing rather than papering over).
*
* RFC-WS-6 session 2.7 see also CLAUDE.md "Database" section
* (byte-stability rule for JSON columns).
*/
final class JsonCanonicalizer
{
/**
* Canonicalize the structure recursively (sort associative keys).
*
* @template T
*
* @param T $value
* @return T
*/
public static function canonicalize(mixed $value): mixed
{
if (! is_array($value)) {
return $value;
}
if ($value === []) {
return $value;
}
if (array_is_list($value)) {
return array_map(self::canonicalize(...), $value);
}
ksort($value);
foreach ($value as $key => $child) {
$value[$key] = self::canonicalize($child);
}
return $value;
}
/**
* Encode a value as canonical JSON.
*
* Use for values stored in MySQL JSON columns where byte-stability
* matters across reads/writes.
*
* @throws \JsonException
*/
public static function encode(mixed $value): string
{
return json_encode(
self::canonicalize($value),
JSON_UNESCAPED_UNICODE | JSON_UNESCAPED_SLASHES | JSON_THROW_ON_ERROR,
);
}
}