knowledgepixels / nanopub-registry / 32111416371

85%

push

github

test: add unit tests for existing functionality (#123)

* test: add unit tests

* refactor: replace System.getenv with Utils.getEnv for environment variable retrieval

* test: add test fixture for nanopublication settings

* test: cover the task state machine and env-gated branches

* refactor: improve readability and efficiency

* test: add unit tests for Utils class to handle null JSON and peer URL exclusions

* test: add parallel failure handling to loadStreamInParallel method

* test: enhance RegistryPeerConnectorTest with additional discoverPubkeys scenarios

* test(deps): add mockito as an agent to maven's surefire plugin

* test(EntryStatus): add additional null value handling and exact spelling checks

* test(RegistryDB): update loadNanopub test to ignore unsigned nanopubs and verify storage behavior

* refactor(Task): remove old session parameter from `runTask` method after merge

* test(Task): follow the task split from #128

Merging main brought in #128, which split every task that cannot be
transactional away from the writes it used to make: INIT_COLLECTIONS now
only does DDL and fetches the intro collection, with the seeding moved to
SEED_TRUST_STATE, and RELEASE_DATA only renames the collections, with the
hash, counter, snapshot and status transition moved to
PUBLISH_TRUST_STATE. The tests still ran only the first half of each pair
while asserting on the second half's effects.

Add a runQueuedSuccessor helper and chain both halves where the
assertions span them, which also gives SEED_TRUST_STATE and
PUBLISH_TRUST_STATE their first coverage.

Assert the exact set of non-transactional tasks rather than just the two
that predate #128, so an unintended opt-out still fails.

releaseDataPublishesTheLoadingCollectionsAndSnapshotsTheState seeds a
toLoad account, which trips the #119 bootstrap hold-back. Set
REGISTRY_PERFORM_FULL_LOAD=false so it exercises the normal publication
path it was written for, and cover the hold-back itself in a new tes... (continued)

941 of 1182 branches covered (79.61%)

Branch coverage included in aggregate %.

2867 of 3300 relevant lines covered (86.88%)

13.47 hits per line