Sy/update metadata validation (#17521)

* add sample tags to validation * changelog * make validation optional * fix tests * lint * add tests and comments * rename tests and comments
DataDog · May 10, 2024 · 98a519e · 98a519e
1 parent a877d95
commit 98a519e
Show file tree

Hide file tree

Showing 4 changed files with 72 additions and 4 deletions.
diff --git a/ddev/changelog.d/17521.added b/ddev/changelog.d/17521.added
@@ -0,0 +1 @@
+Add sample_tags to metadata validation
diff --git a/ddev/src/ddev/cli/validate/metadata.py b/ddev/src/ddev/cli/validate/metadata.py
@@ -130,12 +130,17 @@ def metadata(app: Application, integrations: tuple[str, ...], check_duplicates:
 
                     error_message += f"{current_check.name}:{line} Invalid column {invalid_headers}.\n"
 
-                missing_headers = metadata_utils.ALL_HEADERS.difference(all_keys)
+                missing_headers = metadata_utils.HEADERS_TO_CHECK.difference(all_keys)
                 if missing_headers:
                     errors = True
 
                     error_message += f"{current_check.name}:{line} Missing columns {missing_headers}.\n"
-                continue
+
+                if errors:
+                    # There's now an optional sample_tag column that isn't added yet to the existing metadata.csv
+                    # all_keys will not be same as ALL_HEADERS. But since that sample_tag column is optional and not
+                    # inside HEADERS_TO_CHECK, we should only continue if there's an invalid header or missing_header.
+                    continue
 
             # check duplicate metric name
             duplicate_metric_name = check_duplicate_values(

diff --git a/ddev/src/ddev/cli/validate/metadata_utils.py b/ddev/src/ddev/cli/validate/metadata_utils.py
@@ -5,9 +5,18 @@
 
 REQUIRED_HEADERS = {'metric_name', 'metric_type', 'orientation', 'integration'}
 
-OPTIONAL_HEADERS = {'description', 'interval', 'unit_name', 'per_unit_name', 'short_name', 'curated_metric'}
+OPTIONAL_HEADERS = {
+    'description',
+    'interval',
+    'unit_name',
+    'per_unit_name',
+    'short_name',
+    'curated_metric',
+}
 
-ALL_HEADERS = REQUIRED_HEADERS | OPTIONAL_HEADERS
+EXPERIMENTAL_HEADER = {"sample_tags"}
+ALL_HEADERS = REQUIRED_HEADERS | OPTIONAL_HEADERS | EXPERIMENTAL_HEADER
+HEADERS_TO_CHECK = REQUIRED_HEADERS | OPTIONAL_HEADERS
 
 ORDERED_HEADERS = [
     "metric_name",
@@ -20,6 +29,7 @@
     "integration",
     "short_name",
     "curated_metric",
+    "sample_tags",
 ]
 
 VALID_METRIC_TYPE = {'count', 'gauge', 'rate'}

diff --git a/ddev/tests/cli/validate/test_metrics.py b/ddev/tests/cli/validate/test_metrics.py
@@ -715,6 +715,58 @@ def test_metrics_ordered(fake_repo, ddev, helpers):
     )
 
 
+def test_passing_with_experimental_column(fake_repo, ddev, helpers):
+    # Testing to ensure that experimental header sample_tags is allowed
+    write_file(
+        fake_repo.path / "metadata_integration",
+        'metadata.csv',
+        """metric_name,metric_type,interval,unit_name,per_unit_name,description,orientation,integration,short_name,curated_metric,sample_tags
+metadata_integration.metric_a,gauge,,,,My metric A,0,metadata_integration,,,
+metadata_integration.metric_b,gauge,,,,My metric B,0,metadata_integration,,,
+""",
+    )
+
+    result = ddev('validate', 'metadata', 'metadata_integration')
+
+    assert result.exit_code == 0, result.output
+    assert helpers.remove_trailing_spaces(result.output) == helpers.dedent(
+        """
+        Metrics validation
+
+        Passed: 1
+        """
+    )
+
+
+def test_passing_invalid_experimental_column(fake_repo, ddev, helpers):
+    # Testing to ensure that experimental header sample_tags is allowed. But if other tags are added,
+    # it will be flagged as an error
+    write_file(
+        fake_repo.path / "metadata_integration",
+        'metadata.csv',
+        """metric_name,metric_type,interval,unit_name,per_unit_name,description,orientation,integration,short_name,curated_metric,sample_tags,foo
+metadata_integration.metric_a,gauge,,,,My metric A,0,metadata_integration,,,,
+metadata_integration.metric_b,gauge,,,,My metric B,0,metadata_integration,,,,
+""",
+    )
+    outfile = os.path.join('metadata_integration', 'metadata.csv')
+    result = ddev('validate', 'metadata', 'metadata_integration')
+
+    assert result.exit_code == 1, result.output
+    assert helpers.remove_trailing_spaces(result.output) == helpers.dedent(
+        f"""
+        Metrics validation
+        └── metadata_integration
+            └── {outfile}
+
+                metadata_integration:2 Invalid column {{'foo'}}.
+                metadata_integration:3 Invalid column {{'foo'}}.
+
+        Errors: 1
+        """
+    )
+
+
 def test_metrics_not_ordered(fake_repo, ddev, helpers):
     outfile = os.path.join('metadata_integration', 'metadata.csv')
     result = ddev('validate', 'metadata', 'metadata_integration')