The testListUsers function was using $totalUsers which was removed when
the test was refactored for parallel execution. Now uses $minUsers and
flexible assertions instead of position-based assertions.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Under high parallel load, authentication can transiently fail with 401.
Add retry logic with delays to handle these transient failures.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Update account name before testing to ensure webhook data has
expected 'New Name' value, regardless of test execution order.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Update bucket name before deleting to ensure webhook data has
expected name, regardless of whether testUpdateStorageBucket ran.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- setupDatabase() now fetches existing database on conflict
- setupTable() now fetches existing tables on conflict
- Enables parallel test execution with fixed resource IDs
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- setupDatabase() now fetches existing database on conflict
- setupCollections() now fetches existing collections on conflict
- Enables parallel test execution with fixed resource IDs
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Use assertGreaterThanOrEqual instead of exact count
- Find users by ID instead of assuming position
- Document expected minimum user count for isolated execution
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Create bucket1 within testListBucket instead of relying on
testCreateBucket running first. Accept 409 if bucket already exists.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Use assertGreaterThanOrEqual and find bucket by ID instead of
assuming exact count and position after cursor query.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Account: Add createFreshAccountWithSession() for predictable session/log counts
- Account: Update testGetAccountSessions and testGetAccountLogs to use fresh accounts
- Storage: Fix testListBucket to find bucket by ID instead of assuming first position
- Messaging: Fix testListSubscribers to find subscriber by ID instead of assuming first position
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
These tests were creating resources directly in the shared collections,
which conflicts with setupAttributes() and setupIndexes() when tests
run in parallel. Now each test creates its own dedicated collection
to avoid conflicts.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Adding process ID and random bytes to uniqid() to avoid email
collisions when running tests in parallel with functional mode.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Removed return statements from testManyToOneRelationship and
testManyToManyRelationship which were declared as void. These
cause PHP fatal errors in PHP 8.4+.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The testUpdateDocument method was declared as void but returned [].
This causes a PHP fatal error in PHP 8.4.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add search filters to isolate test data in testListSites to prevent
count assertions from failing when other test methods create sites
during parallel execution.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Added setupDatabase() helper with caching to all permission test files
- Removed @depends from LegacyPermissionsMemberTest, LegacyPermissionsTeamTest
- Removed @depends from TablesDBPermissionsMemberTest, TablesDBPermissionsTeamTest
- Partial GraphQL test updates from agents
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Enables --functional mode for parallel test execution by:
- Adding helper methods for independent test setup
- Removing all #[Depends] annotations
- Using static caching for shared resources
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Use /databases endpoint instead of /tablesdb for database creation
- Use /columns/string instead of /attributes/string for table columns
- Fix testAttributesCollectionsAPI return type from array to void
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add helper methods createCollectionWithAttribute(), createTableWithAttribute(),
createCollectionWithIndex(), createTableWithIndex() to RealtimeConsoleClientTest
- Add createTeam() helper to RealtimeCustomClientTest
- Replace all @depends annotations with direct helper method calls
- Enable tests to run in parallel with --functional flag
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Fix typo CLient -> Client in DatabasesBase.php:4568
- Fix typo CLient -> Client in StorageCustomClientTest.php:428
- Fix cleanupFunction() call with extra parameter in FunctionsScheduleTest.php:175
- Increase message scheduling test timeouts from 30s to 180s for CI stability
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The root cause of flaky attribute processing tests is that a single
appwrite-worker-databases container cannot keep up with attribute
creation jobs when tests run in parallel via paratest.
This fix:
- Scales the databases worker to 4 instances during E2E service tests
- Reverts timeout increases that masked the underlying issue
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Increase all default timeouts in SchemaPolling trait from 5 minutes to
10 minutes. With tests running in parallel via paratest, the attribute
processing worker can become backlogged under heavy CI load, requiring
longer wait times for attributes to transition from 'processing' to
'available' status.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Increase all default timeouts in SchemaPolling trait from 3 minutes to
5 minutes to handle slower attribute processing when running tests in
parallel under heavy CI load.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The PNG file size varies depending on ImageMagick version and
compression settings. The meaningful validations (dimensions,
format, content-type) are already being tested.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Fix OR precheck bug: skip attribute existence check when OR conditions
exist, since OR can match with partial attributes present
- Remove dead code in app/realtime.php (unused $names and $channels variables)
- Add tests for OR queries with missing attributes in one branch
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
RuntimeQuery::filter() now returns null when the query doesn't match,
instead of an empty array. This distinguishes between "no match" and
"match with empty payload", fixing the issue where subscriptions with
empty payloads weren't being delivered.
Updated Realtime::getSubscribers() to check for null instead of using
!empty(), and updated all tests to expect null for non-matches.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add missing isSelectAll() method to RuntimeQuery class
- Update RuntimeQueryTest to use compile() before filter() since filter()
expects pre-compiled query arrays, not Query objects
- Remove incorrect break statement in Realtime::getSubscribers() that was
stopping iteration after the first matching channel, causing subscribers
to not be found for subsequent channels
- Use local variable $subscriptionsByChannel to avoid unused variable warning
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Increase all default timeouts in SchemaPolling trait from 2 minutes to
3 minutes to handle slower attribute processing in CI environments.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Remove explicit sleep and increase assertEventually timeout from 60s to
180s to account for slower deployment and scheduling in CI. Also use
assertGreaterThanOrEqual to handle cases where multiple executions
may have run.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Skip FunctionsCustomServerTest::testCreateFunctionAndDeploymentFromTemplateCommit when GitHub API is rate-limited or unavailable
- Increase waitForAllAttributes timeout to 3 minutes for testBulkDeleteMatchingUpdatedDocuments to handle slow attribute processing in CI
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
