clearer txt report error indicators (#139)

* more meaningful api names

* clearer txt report error indicators

api/trace.go

@@ -25,9 +25,9 @@ type Tracer interface {
 // Trace a single time trace
 type Trace interface {
 	ID() string
-	Elapsed() time.Duration
-	System() time.Duration
-	User() time.Duration
+	PerceivedTime() time.Duration
+	SystemCPUTime() time.Duration
+	UserCPUTime() time.Duration
 	Error() error
 }
 

internal/report/csv_rawdata_report_writer.go

@@ -50,9 +50,9 @@ func (rw CsvStreamReportWriter) Handle(trace api.Trace) (err error) {
 		timeStr,
 		trace.ID(),
 		strings.Join(rw.ctx.Labels, ","),
-		fmt.Sprintf("%d", trace.Elapsed()),
-		fmt.Sprintf("%d", trace.User()),
-		fmt.Sprintf("%d", trace.System()),
+		fmt.Sprintf("%d", trace.PerceivedTime()),
+		fmt.Sprintf("%d", trace.UserCPUTime()),
+		fmt.Sprintf("%d", trace.SystemCPUTime()),
 		fmt.Sprintf("%v", trace.Error() != nil),
 	})
 

internal/report/md_rawdata_report_writer.go

@@ -38,9 +38,9 @@ func (rw MarkdownStreamReportWriter) Handle(trace api.Trace) (err error) {
 		timeStr,
 		trace.ID(),
 		strings.Join(rw.ctx.Labels, ","),
-		fmt.Sprintf("%d", trace.Elapsed()),
-		fmt.Sprintf("%d", trace.User()),
-		fmt.Sprintf("%d", trace.System()),
+		fmt.Sprintf("%d", trace.PerceivedTime()),
+		fmt.Sprintf("%d", trace.UserCPUTime()),
+		fmt.Sprintf("%d", trace.SystemCPUTime()),
 		fmt.Sprintf("%v", trace.Error() != nil),
 	})
 

internal/report/raw_report_test_utils.go

@@ -28,9 +28,9 @@ func assertRawTraceRecord(t *testing.T, trace api.Trace, actualRecord []string)
 	assert.NoError(t, err)
 	assert.Equal(t, trace.ID(), actualRecord[1])
 	assert.Equal(t, expectedLabels, actualRecord[2])
-	assert.Equal(t, fmt.Sprint(trace.Elapsed().Nanoseconds()), actualRecord[3])
-	assert.Equal(t, fmt.Sprint(trace.User().Nanoseconds()), actualRecord[4])
-	assert.Equal(t, fmt.Sprint(trace.System().Nanoseconds()), actualRecord[5])
+	assert.Equal(t, fmt.Sprint(trace.PerceivedTime().Nanoseconds()), actualRecord[3])
+	assert.Equal(t, fmt.Sprint(trace.UserCPUTime().Nanoseconds()), actualRecord[4])
+	assert.Equal(t, fmt.Sprint(trace.SystemCPUTime().Nanoseconds()), actualRecord[5])
 	assert.Equal(t, fmt.Sprint(trace.Error() != nil), actualRecord[6])
 }
 

internal/report/text_report_writer.go

@@ -12,6 +12,8 @@ import (
 	log "github.com/sirupsen/logrus"
 )
 
+const attentionIndicatorRune = '•'
+
 // textReportWriter a simple human readable test report writer
 type textReportWriter struct {
 	writer  *bufio.Writer
@@ -102,6 +104,7 @@ func (trw textReportWriter) Write(summary api.Summary, config api.BenchmarkSpec,
 
 		trw.writeDurationProperty("user", trw.hiblue, userStats.Mean)
 		trw.writeDurationProperty("system", trw.hiblue, sysStats.Mean)
+
 		trw.writeErrorRateStat("errors", stats.ErrorRate)
 
 		trw.writeNewLine()
@@ -144,14 +147,15 @@ func (trw textReportWriter) writeDurationProperty(name string, c *color.Color, f
 	trw.writeProperty(name, FormatReportDuration(f), c)
 }
 
-func (trw textReportWriter) writeErrorRateStat(name string, f func() float64) {
-	value := f()
-	errorRate := int(value * 100)
-	if errorRate > 0 {
-		trw.writeString(trw.yellow.Sprintf("%11s: %d%%", name, errorRate))
-	} else {
-		trw.writeString(fmt.Sprintf("%11s: %d%%", name, errorRate))
+func (trw textReportWriter) writeErrorRateStat(name string, errorRate func() float64) {
+	errorRatePercent := int(errorRate() * 100)
+	var attentionIndicator = ""
+
+	if errorRatePercent > 10 {
+		attentionIndicator = trw.red.Sprintf("%c", attentionIndicatorRune)
 	}
+
+	trw.writeString(fmt.Sprintf("%11s: %d%% %s", name, errorRatePercent, attentionIndicator))
 }
 
 func (trw textReportWriter) writeInt64StatLine(name string, f func() (int64, error)) {

internal/report/text_report_writer_test.go

@@ -67,7 +67,7 @@ func testTxtSanity(t *testing.T, colorsOn bool) {
 	assert.Equal(t, 1, strings.Count(text, "mean: 2.0s"))
 
 	assert.Equal(t, 1, strings.Count(text, "errors: 0%"))
-	assert.Equal(t, 1, strings.Count(text, "errors: 100%"))
+	assert.Equal(t, 1, strings.Count(text, "errors: 100% •"))
 
 	assert.Equal(t, 1, strings.Count(text, "system: 1"))
 	assert.Equal(t, 1, strings.Count(text, "system: 2"))

pkg/command_exec.go

@@ -34,16 +34,16 @@ func (ce *commandExecutor) ExecuteFn(cmdSpec *api.CommandSpec, defaultWorkingDir
 
 	return func() (execInfo *api.ExecutionInfo, err error) {
 		defer cancel()
-		
+
 		startTime := time.Now()
 		err = execCmd.Run()
 		perceivedTime := time.Since(startTime)
 		state := execCmd.ProcessState
 		if state != nil && state.Exited() {
 			execInfo = &api.ExecutionInfo{
-				ExitCode:   state.ExitCode(),
-				UserTime:   state.UserTime(),
-				SystemTime: state.SystemTime(),
+				ExitCode:      state.ExitCode(),
+				UserTime:      state.UserTime(),
+				SystemTime:    state.SystemTime(),
 				PerceivedTime: perceivedTime,
 			}
 

pkg/fake_summary.go

@@ -11,8 +11,8 @@ func NewFakeTrace(id string, elapsed, userTime, sysTime time.Duration, err error
 	return &trace{
 		id:            id,
 		perceivedTime: elapsed,
-		usrTime:       userTime,
-		sysTime:       sysTime,
+		usrCPUTime:    userTime,
+		sysCPUTime:    sysTime,
 		error:         err,
 	}
 }

pkg/summary.go

@@ -23,9 +23,9 @@ func NewSummary(tracesByID map[api.ID][]api.Trace) api.Summary {
 		errorCount := 0
 
 		for ti := range traces {
-			perceivedSamples[ti] = float64(traces[ti].Elapsed().Nanoseconds())
-			systemSamples[ti] = float64(traces[ti].System().Nanoseconds())
-			userSamples[ti] = float64(traces[ti].User().Nanoseconds())
+			perceivedSamples[ti] = float64(traces[ti].PerceivedTime().Nanoseconds())
+			systemSamples[ti] = float64(traces[ti].SystemCPUTime().Nanoseconds())
+			userSamples[ti] = float64(traces[ti].UserCPUTime().Nanoseconds())
 			if traces[ti].Error() != nil {
 				errorCount++
 			}

pkg/trace.go

@@ -13,25 +13,25 @@ type tracer struct {
 type trace struct {
 	id            string
 	perceivedTime time.Duration
-	usrTime       time.Duration
-	sysTime       time.Duration
+	usrCPUTime    time.Duration
+	sysCPUTime    time.Duration
 	error         error
 }
 
 func (t trace) ID() string {
 	return t.id
 }
 
-func (t trace) Elapsed() time.Duration {
+func (t trace) PerceivedTime() time.Duration {
 	return t.perceivedTime
 }
 
-func (t trace) System() time.Duration {
-	return t.sysTime
+func (t trace) SystemCPUTime() time.Duration {
+	return t.sysCPUTime
 }
 
-func (t trace) User() time.Duration {
-	return t.usrTime
+func (t trace) UserCPUTime() time.Duration {
+	return t.usrCPUTime
 }
 
 func (t trace) Error() error {
@@ -60,7 +60,7 @@ func (tr *tracer) Start(i api.Identifiable) api.End {
 func (tr *tracer) endFn(t trace) api.End {
 	return func(execInfo *api.ExecutionInfo, exitError error) {
 		if execInfo != nil {
-			t.perceivedTime, t.usrTime, t.sysTime = execInfo.PerceivedTime, execInfo.UserTime, execInfo.SystemTime
+			t.perceivedTime, t.usrCPUTime, t.sysCPUTime = execInfo.PerceivedTime, execInfo.UserTime, execInfo.SystemTime
 		}
 		t.error = exitError
 

pkg/trace_test.go

@@ -32,9 +32,9 @@ func Test_tracer_endFn(t *testing.T) {
 	)
 	received := <-tracer.Stream()
 
-	assert.Equal(t, expectedPerceivedTime, received.Elapsed())
-	assert.Equal(t, expectedUserTime, received.User())
-	assert.Equal(t, expectedSysTime, received.System())
+	assert.Equal(t, expectedPerceivedTime, received.PerceivedTime())
+	assert.Equal(t, expectedUserTime, received.UserCPUTime())
+	assert.Equal(t, expectedSysTime, received.SystemCPUTime())
 	assert.Equal(t, expectedError, received.Error())
 	assert.Equal(t, expectedID, received.ID())
 }
@@ -53,9 +53,9 @@ func Test_tracer_endFn_withError(t *testing.T) {
 	)
 	received := <-tracer.Stream()
 
-	assert.Equal(t, expectedDuration, received.Elapsed())
-	assert.Equal(t, expectedDuration, received.User())
-	assert.Equal(t, expectedDuration, received.System())
+	assert.Equal(t, expectedDuration, received.PerceivedTime())
+	assert.Equal(t, expectedDuration, received.UserCPUTime())
+	assert.Equal(t, expectedDuration, received.SystemCPUTime())
 	assert.Equal(t, expectedError, received.Error())
 	assert.Equal(t, expectedID, received.ID())
 }