"""Tests for background jobs storage helpers.""" import time import uuid import pytest from gn_libs import sqlite3 from gn_libs.jobs import jobs from gn_libs.jobs.migrations import run_migrations @pytest.fixture() def jobsdb(tmp_path): """Create a migrated jobs database.""" db = tmp_path / "jobs.db" run_migrations(str(db)) return db def _job(conn, job_type, **metadata): """Create a job with the given type and metadata.""" return jobs.initialise_job( conn, str(uuid.uuid4()), ["true"], job_type, **metadata) @pytest.mark.unit_test def test_jobs_metadata_has_key_value_index(jobsdb): """ GIVEN: a migrated jobs database WHEN: the schema is inspected THEN: the metadata key/value index exists """ with sqlite3.connection(str(jobsdb)) as conn, sqlite3.cursor(conn) as cursor: cursor.execute( "SELECT name FROM sqlite_master WHERE type='index' " "AND name='idx_tbl_jobs_metadata_cols_key_value'") assert cursor.fetchone() is not None @pytest.mark.unit_test def test_jobs_by_metadata_matches_all_pairs(jobsdb): """ GIVEN: jobs with various metadata WHEN: jobs are looked up by several metadata pairs THEN: only jobs matching all pairs are returned, with their metadata """ with sqlite3.connection(str(jobsdb)) as conn: wanted = _job(conn, "qc", genotype_collector_job_id="c1") _job(conn, "qc", genotype_collector_job_id="c2") _job(conn, "other", genotype_collector_job_id="c1") found = jobs.jobs_by_metadata( conn, {"genotype_collector_job_id": "c1", "job-type": "qc"}) assert [job["job_id"] for job in found] == [wanted["job_id"]] assert found[0]["metadata"]["genotype_collector_job_id"] == "c1" @pytest.mark.unit_test def test_jobs_by_metadata_returns_newest_first(jobsdb): """ GIVEN: two matching jobs created at different times WHEN: jobs are looked up by metadata THEN: the newest job is returned first """ with sqlite3.connection(str(jobsdb)) as conn: first = _job(conn, "qc", genotype_collector_job_id="c1") time.sleep(0.01) second = _job(conn, "qc", genotype_collector_job_id="c1") found = jobs.jobs_by_metadata( conn, {"genotype_collector_job_id": "c1"}) assert [job["job_id"] for job in found] == [ second["job_id"], first["job_id"]] @pytest.mark.unit_test def test_jobs_by_metadata_returns_empty_when_no_match(jobsdb): """ GIVEN: jobs that do not match the query WHEN: jobs are looked up by metadata THEN: an empty tuple is returned """ with sqlite3.connection(str(jobsdb)) as conn: _job(conn, "qc", genotype_collector_job_id="c1") assert jobs.jobs_by_metadata( conn, {"genotype_collector_job_id": "nope"}) == () @pytest.mark.unit_test def test_initialise_job_treats_metadata_kwargs_as_metadata(jobsdb): """ GIVEN: metadata and reserved operational arguments WHEN: a job is initialised THEN: the metadata is stored, while expiry_seconds is not metadata """ with sqlite3.connection(str(jobsdb)) as conn: job = jobs.initialise_job( conn, str(uuid.uuid4()), ["true"], "qc", genotype_collector_job_id="c1", external_id="user-1", expiry_seconds=60) stored = jobs.job(conn, job["job_id"]) assert stored["metadata"]["genotype_collector_job_id"] == "c1" assert stored["metadata"]["external_id"] == "user-1" assert "expiry_seconds" not in stored["metadata"] assert jobs.jobs_by_external_id(conn, "user-1")[0]["job_id"] == ( job["job_id"])