fix(cli): fix example data contract yaml + update airflow codecov (#9707)

.github/workflows/airflow-plugin.yml

@@ -87,8 +87,8 @@ jobs:
           token: ${{ secrets.CODECOV_TOKEN }}
           directory: .
           fail_ci_if_error: false
-          flags: airflow-${{ matrix.python-version }}-${{ matrix.extraPythonRequirement }}
-          name: pytest-airflow
+          flags: airflow,airflow-${{ matrix.extra_pip_extras }}
+          name: pytest-airflow-${{ matrix.python-version }}-${{ matrix.extra_pip_requirements }}
           verbose: true
 
   event-file:

metadata-ingestion-modules/airflow-plugin/build.gradle

@@ -108,7 +108,7 @@ task testQuick(type: Exec, dependsOn: installDevTest) {
   inputs.files(project.fileTree(dir: "src/", include: "**/*.py"))
   inputs.files(project.fileTree(dir: "tests/"))
   commandLine 'bash', '-x', '-c',
-    "source ${venv_name}/bin/activate && pytest  -vv --continue-on-collection-errors --junit-xml=junit.quick.xml"
+    "source ${venv_name}/bin/activate && pytest --cov-config=setup.cfg --cov-report xml:coverage_quick.xml -vv --continue-on-collection-errors --junit-xml=junit.quick.xml"
 }
 
 

metadata-ingestion-modules/airflow-plugin/tests/conftest.py

@@ -1,6 +1,17 @@
+import pathlib
+import site
+
+
 def pytest_addoption(parser):
     parser.addoption(
         "--update-golden-files",
         action="store_true",
         default=False,
     )
+
+
+# See https://coverage.readthedocs.io/en/latest/subprocess.html#configuring-python-for-sub-process-measurement
+coverage_startup_code = "import coverage; coverage.process_startup()"
+site_packages_dir = pathlib.Path(site.getsitepackages()[0])
+pth_file_path = site_packages_dir / "datahub_coverage_startup.pth"
+pth_file_path.write_text(coverage_startup_code)

metadata-ingestion/examples/data_contract/pet_of_the_week.dhub.dc.yaml

@@ -1,21 +1,29 @@
-# id: pet_details_dc # Optional: This is the unique identifier for the data contract
-display_name: Data Contract for SampleHiveDataset
+version: 1  # datahub yaml format version
+
+# Note: this data contract yaml format is still in development, and will likely
+# change in backwards-incompatible ways in the future.
+
 entity: urn:li:dataset:(urn:li:dataPlatform:hive,SampleHiveDataset,PROD)
 freshness:
-  time: 0700
-  granularity: DAILY
+  type: cron
+  cron: 0 7 * * *  # 7am daily
+  timezone: America/Los_Angeles
 schema:
-  properties:
-    field_foo:
-      type: string
-      native_type: VARCHAR(100)
-    field_bar:
-      type: boolean
-  required:
-    - field_bar
+  type: json-schema
+  json-schema:
+    properties:
+      field_foo:
+        type: string
+        native_type: VARCHAR(100)
+      field_bar:
+        type: boolean
+    required:
+      - field_bar
 data_quality:
-  - type: column_range
-    config:
-      column: field_foo
-      min: 0
-      max: 100
+  - type: unique
+    column: field_foo
+  - type: custom_sql
+    sql: SELECT COUNT(*) FROM SampleHiveDataset
+    operator:
+      type: greater_than
+      value: 100