Fragment LC tests

.github/workflows/test.yaml

@@ -30,7 +30,7 @@ jobs:
         timeout-minutes: 20
         run: |
           PYTHONPATH=$PWD/gnomad_methods:$PWD/seqr-loading-pipelines \
-          coverage run -m pytest test --junitxml=test-execution.xml
+          coverage run -m pytest -n auto test --junitxml=test-execution.xml
 
           rc=$?
           coverage xml

.pre-commit-config.yaml

@@ -13,7 +13,7 @@ repos:
       - id: detect-private-key
       - id: debug-statements
       - id: check-added-large-files
-        exclude: '\.*.interval_list'
+        exclude: '\.*.interval_list|test/data/large_cohort/compressed_dirs/.*'
 
   - repo: https://github.com/populationgenomics/pre-commits
     rev: "v0.1.3"

cpg_workflows/jobs/vep.py

@@ -144,7 +144,6 @@ def vep_one(
     vep_image = image_path(f'vep_{vep_version}')
     vep_mount_path = to_path(reference_path(f'vep_{vep_version}_mount'))
     assert all([vep_image, vep_mount_path])
-    logging.info(f'Using VEP {vep_version}')
 
     j = b.new_job('VEP', (job_attrs or {}) | dict(tool=f'VEP {vep_version}'))
     j.image(vep_image)

cpg_workflows/large_cohort/dense_subset.py

@@ -7,10 +7,7 @@
 from cpg_workflows.utils import can_reuse
 
 
-def run(
-    vds_path: str,
-    out_dense_mt_path: str,
-) -> hl.MatrixTable:
+def run(vds_path: str, out_dense_mt_path: str) -> hl.MatrixTable:
     """
     Filter a sparse VariantDataset to a set of predetermined QC sites
     and return a dense MatrixTable with split multiallelics.

requirements-dev.txt

@@ -2,8 +2,9 @@ pre-commit
 pylint
 bump2version
 black
-pytest
+pytest>8
 pytest_mock
+pytest-xdist>=3.6.0
 mypy
 cpg-utils>=5.0.11
 tenacity

setup.py

@@ -13,7 +13,7 @@
     license='MIT',
     packages=find_packages(),
     install_requires=[
-        'cpg-utils>=5.0.11',
+        'cpg-utils>=5.1.1',
         'cyvcf2==0.30.18',
         'analysis-runner>=2.43.3',
         'hail==0.2.132',  # Pin Hail at CPG's installed version
@@ -40,6 +40,7 @@
     extras_require={
         'test': [
             'pytest',
+            'pytest-xdist',
             'pytest-mock',
             'coverage',
         ],

test/conftest.py

@@ -10,10 +10,19 @@
 import cpg_workflows.metamist
 import cpg_workflows.stages.gatk_sv
 import cpg_workflows.workflow
+from cpg_utils.hail_batch import start_query_context
+
+
+@pytest.fixture(autouse=True, scope='session')
+def start_up_query():
+    """Start up the query backend once per session."""
+    start_query_context()
+    yield
 
 
 @pytest.fixture(autouse=True, scope='function')
 def pre_and_post_test():
+
     # Set a dummy google cloud project to avoid errors when running tests for tests
     # that use the google cloud.
     with mock.patch.dict(