[processor/tailsamplingprocessor] and support for string invert match (open-telemetry#9768)

* [processor/tailsamplingprocessor] and support for string invert

* Add open-telemetry#9553 to changelog.

Author: tillig

CHANGELOG.md

@@ -10,6 +10,8 @@
 
 ### 💡 Enhancements 💡
 
+- `tailsamplingprocessor`: Add support for string invert matching to `and` policy (#9553)
+
 ### 🧰 Bug fixes 🧰
 
 ## v0.52.0
@@ -63,7 +65,7 @@
 ### 🚩 Deprecations 🚩
 
 - `exporter/azuremonitor`: Deprecate use of LogRecord.Name as the log envelope category name. There is no replacement. (#9258)
-- `processor/k8sattributes`: Deprecate use of k8s.cluster.name metadata parameter (obsolete) (#9968) 
+- `processor/k8sattributes`: Deprecate use of k8s.cluster.name metadata parameter (obsolete) (#9968)
 
 ### 🚀 New components 🚀
 
@@ -75,7 +77,7 @@
 ### 💡 Enhancements 💡
 
 - `cmd/mdatagen`: Replace enum attributes values with typed constants (#9683)
-- `elasticsearchreceiver`: Update metrics scope name from `otelcol/elasticsearch` 
+- `elasticsearchreceiver`: Update metrics scope name from `otelcol/elasticsearch`
   to `otelcol/elasticsearchreceiver` (#9757)
 - `k8sclusterreceiver`: Validate that k8s API supports a resource before setting up a watcher for it (#9523)
 - `internal/stanza`: Add support for `remove` operator (#9524)
@@ -161,7 +163,7 @@
 
 This release contains an issue in
 [Prometheus receiver](https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/receiver/prometheusreceiver)
-causing 30% memory consumption increase when there is a lot of target churn. The issue is currently being 
+causing 30% memory consumption increase when there is a lot of target churn. The issue is currently being
 investigated and will be fixed in one of the new releases. More details:
 https://github.com/open-telemetry/opentelemetry-collector-contrib/issues/9278.
 
@@ -176,7 +178,7 @@ https://github.com/open-telemetry/opentelemetry-collector-contrib/issues/9278.
 - `dynatraceexporter`: Make `serialization` package `/internal` (#9097)
 - `attributesprocessor`: Remove log names from filters (#9131)
 - `k8sclusterreceiver`: The `receiver.k8sclusterreceiver.reportCpuMetricsAsDouble` feature gate is now enabled by default (#9367)
-  - Users may have to update monitoring for a few Kubernetes cpu metrics, for 
+  - Users may have to update monitoring for a few Kubernetes cpu metrics, for
     more details see [feature-gate-configurations](https://github.com/open-telemetry/opentelemetry-collector-contrib/tree/main/receiver/k8sclusterreceiver#feature-gate-configurations).
 
 ### 🚩 Deprecations 🚩

processor/tailsamplingprocessor/internal/sampling/and.go

@@ -45,7 +45,7 @@ func (c *And) Evaluate(traceID pcommon.TraceID, trace *TraceData) (Decision, err
 		if err != nil {
 			return Unspecified, err
 		}
-		if decision == NotSampled {
+		if decision == NotSampled || decision == InvertNotSampled {
 			return NotSampled, nil
 		}
 

processor/tailsamplingprocessor/internal/sampling/and_test.go

@@ -78,3 +78,59 @@ func TestAndEvaluatorSampled(t *testing.T) {
 	assert.Equal(t, decision, Sampled)
 
 }
+
+func TestAndEvaluatorStringInvertSampled(t *testing.T) {
+	n1 := NewStringAttributeFilter(zap.NewNop(), "attribute_name", []string{"no_match"}, false, 0, true)
+	n2, err := NewStatusCodeFilter(zap.NewNop(), []string{"ERROR"})
+	if err != nil {
+		t.FailNow()
+	}
+
+	and := NewAnd(zap.NewNop(), []PolicyEvaluator{n1, n2})
+
+	traces := ptrace.NewTraces()
+	rs := traces.ResourceSpans().AppendEmpty()
+	ils := rs.ScopeSpans().AppendEmpty()
+
+	span := ils.Spans().AppendEmpty()
+	span.Attributes().InsertString("attribute_name", "attribute_value")
+	span.Status().SetCode(ptrace.StatusCodeError)
+	span.SetTraceID(pcommon.NewTraceID([16]byte{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16}))
+	span.SetSpanID(pcommon.NewSpanID([8]byte{1, 2, 3, 4, 5, 6, 7, 8}))
+
+	trace := &TraceData{
+		ReceivedBatches: []ptrace.Traces{traces},
+	}
+	decision, err := and.Evaluate(traceID, trace)
+	require.NoError(t, err, "Failed to evaluate and policy: %v", err)
+	assert.Equal(t, decision, Sampled)
+
+}
+
+func TestAndEvaluatorStringInvertNotSampled(t *testing.T) {
+	n1 := NewStringAttributeFilter(zap.NewNop(), "attribute_name", []string{"attribute_value"}, false, 0, true)
+	n2, err := NewStatusCodeFilter(zap.NewNop(), []string{"ERROR"})
+	if err != nil {
+		t.FailNow()
+	}
+
+	and := NewAnd(zap.NewNop(), []PolicyEvaluator{n1, n2})
+
+	traces := ptrace.NewTraces()
+	rs := traces.ResourceSpans().AppendEmpty()
+	ils := rs.ScopeSpans().AppendEmpty()
+
+	span := ils.Spans().AppendEmpty()
+	span.Attributes().InsertString("attribute_name", "attribute_value")
+	span.Status().SetCode(ptrace.StatusCodeError)
+	span.SetTraceID(pcommon.NewTraceID([16]byte{1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16}))
+	span.SetSpanID(pcommon.NewSpanID([8]byte{1, 2, 3, 4, 5, 6, 7, 8}))
+
+	trace := &TraceData{
+		ReceivedBatches: []ptrace.Traces{traces},
+	}
+	decision, err := and.Evaluate(traceID, trace)
+	require.NoError(t, err, "Failed to evaluate and policy: %v", err)
+	assert.Equal(t, decision, NotSampled)
+
+}