feat: Split +Inf overflow alerts by label-set (#5)

* Split +Inf overflow alerts by label-set; show metric description
* Improve the message wording to not assume that vales are durations
* Add histogram unit guesser
* Add changelog and bump version to 005
* Fix e2e tests. Ensure CI runs them

Author: vikin91

.github/workflows/ci.yml

@@ -37,15 +37,32 @@ jobs:
             exit 1
           fi
 
-      - name: Run tests
-        run: go test -v -race -coverprofile=coverage.out ./...
+      - name: Run unit tests
+        run: go test -v -race -coverprofile=coverage.out ./internal/...
 
       - name: Upload coverage
         uses: codecov/codecov-action@v4
         with:
           files: coverage.out
           fail_ci_if_error: false
 
+  e2e:
+    runs-on: ubuntu-latest
+    needs: build
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Set up Go
+        uses: actions/setup-go@v5
+        with:
+          go-version: '1.21'
+
+      - name: Build binary
+        run: go build -o bin/metrics-analyzer ./cmd/metrics-analyzer
+
+      - name: Run E2E tests
+        run: go test -v ./cmd/...
+
   validate-rules:
     runs-on: ubuntu-latest
     needs: build

CHANGELOG.md

@@ -0,0 +1,34 @@
+# Changelog
+
+All notable changes to this project are documented in this file.
+
+## 0.0.5
+
+- Split histogram `+Inf` overflow reporting by label-set so problematic series are listed individually.
+- Added metric description (Prometheus HELP text) to overflow findings.
+- Improved overflow wording and added on-the-fly unit guessing for histogram outputs.
+- Improved release/docs workflow (release target, GitHub Pages docs, container registry update).
+
+## 0.0.4
+
+- Added containerization and Helm chart support for deployment.
+- Fixed Helm chart packaging/versioning issues for `0.0.4`.
+- Continued rule quality improvements and review metadata propagation.
+
+## 0.0.3
+
+- Introduced rule review metadata and broader rule/test polishing.
+- Added/expanded tests for gauge-threshold rules.
+- Improved readability of numeric output formatting in reports.
+
+## 0.0.2
+
+- Added web server mode for serving analyzer output.
+
+## 0.0.1
+
+- Initial usable release with the results of the AI hackathon, including:
+  - interactive TUI mode,
+  - demo assets,
+  - generic histogram rule support,
+  - first web-server iteration.

VERSION

@@ -1 +1 @@
-0.0.4
+0.0.5

cmd/metrics-analyzer/main_test.go

@@ -28,6 +28,7 @@ func TestE2EAnalyzeCommand(t *testing.T) {
 				"analyze",
 				"--rules", "../../testdata/fixtures",
 				"--format", "markdown",
+				"--template", "../../templates/markdown.tmpl",
 				"--output", "/tmp/test_e2e_report.md",
 				"../../testdata/fixtures/sample_metrics.txt",
 			},
@@ -209,9 +210,12 @@ func TestE2EFullWorkflow(t *testing.T) {
 	metricsPath := filepath.Join("..", "..", "testdata", "fixtures", "sample_metrics.txt")
 	rulesPath := filepath.Join("..", "..", "testdata", "fixtures")
 
+	templatePath := filepath.Join("..", "..", "templates", "markdown.tmpl")
+
 	cmd := exec.Command(absPath, "analyze",
 		"--rules", rulesPath,
 		"--format", "markdown",
+		"--template", templatePath,
 		"--output", "/tmp/e2e_test_report.md",
 		metricsPath,
 	)

internal/evaluator/evaluator_common.go

@@ -155,7 +155,7 @@ func EvaluateAllRules(rulesList []rules.Rule, metrics parser.MetricsData, loadLe
 	}
 
 	// Apply general histogram +Inf overflow rule to all histogram metrics
-	infOverflowResults := EvaluateHistogramInfOverflow(metrics, loadLevel)
+	infOverflowResults := EvaluateHistogramInfOverflow(metrics)
 	for _, result := range infOverflowResults {
 		report.Results = append(report.Results, result)