Updated the normalize command to correct common errors in WhisperX-generated input transcripts

README.md

@@ -142,16 +142,16 @@ Global flags:
 - Top-level object with a `segments` array.
 - Bare top-level array of segment objects (for example, Audita-style output).
 
-`normalize` required segment fields:
-
-- `start`
-- `end`
-- `speaker`
-- `text`
-
 `normalize` behavior:
 
-- Validates `start >= 0`, `end >= start`, and non-empty `speaker`.
+- Repairs missing timing fields deterministically:
+  if one of `start`/`end` is present, sets both to that value;
+  if both are missing, uses midpoint of previous `end` and next `start`,
+  with edge fallback to available neighbor and `0.0` for single-segment inputs.
+- If `end < start`, swaps them.
+- Fills missing/empty `speaker` with `Unknown_Speaker`.
+- Drops segments with missing, empty, or whitespace-only `text`.
+- Validates repaired timing with `start >= 0`.
 - Accepts existing input `id` values as provenance only.
 - Reassigns output segment IDs sequentially from `1` to `N`.
 - Sorts deterministically by `(start, end, original_input_index, speaker)`.

architecture.md

@@ -223,7 +223,7 @@ Design constraints:
 
 - `normalize` runs outside the merge pipeline and does not invoke merge preprocessing or postprocessing modules.
 - `normalize` accepts two input shapes: object-with-`segments` and bare segment arrays.
-- `normalize` validates required segment fields (`start`, `end`, `speaker`, `text`) and timing/speaker constraints.
+- `normalize` applies deterministic repair rules for missing/irregular `start`, `end`, and `speaker`, and drops segments with missing/empty `text`.
 - `normalize` sorts segments deterministically by chronological keys and stable input-index tie-breakers.
 - `normalize` assigns fresh sequential output IDs (`1..N`) after sorting.
 - `normalize` validates final output against the selected schema before writing.

internal/cli/normalize_test.go

@@ -273,6 +273,66 @@ func TestNormalizeEmptySegmentsArrayProducesValidOutput(t *testing.T) {
 	}
 }
 
+func TestNormalizeRepairsAndDropsDefectiveSegments(t *testing.T) {
+	dir := t.TempDir()
+	input := writeJSONFile(t, dir, "input.json", `[
+		{"start": 5, "speaker": "", "text": "keep-a"},
+		{"end": 3, "speaker": "  ", "text": "keep-b"},
+		{"speaker": "A"},
+		{"speaker": "A", "text": "   "},
+		{"start": 9, "end": 4, "speaker": "B", "text": "keep-c"}
+	]`)
+	output := filepath.Join(dir, "normalized.json")
+	reportPath := filepath.Join(dir, "report.json")
+
+	err := executeNormalize(
+		"--input-file", input,
+		"--output-file", output,
+		"--report-file", reportPath,
+		"--output-schema", config.OutputSchemaIntermediate,
+	)
+	if err != nil {
+		t.Fatalf("normalize failed: %v", err)
+	}
+
+	var transcript schema.IntermediateTranscript
+	readJSON(t, output, &transcript)
+	if len(transcript.Segments) != 3 {
+		t.Fatalf("segment count = %d, want 3", len(transcript.Segments))
+	}
+	if transcript.Segments[0].Start != 3 || transcript.Segments[0].End != 3 {
+		t.Fatalf("segment[0] timing = %v..%v, want 3..3", transcript.Segments[0].Start, transcript.Segments[0].End)
+	}
+	if transcript.Segments[1].Start != 4 || transcript.Segments[1].End != 9 {
+		t.Fatalf("segment[1] timing = %v..%v, want 4..9", transcript.Segments[1].Start, transcript.Segments[1].End)
+	}
+	if transcript.Segments[2].Start != 5 || transcript.Segments[2].End != 5 {
+		t.Fatalf("segment[2] timing = %v..%v, want 5..5", transcript.Segments[2].Start, transcript.Segments[2].End)
+	}
+	if transcript.Segments[0].Speaker != "Unknown_Speaker" || transcript.Segments[2].Speaker != "Unknown_Speaker" {
+		t.Fatalf("expected Unknown_Speaker placeholders, got %#v", transcript.Segments)
+	}
+
+	var rpt report.Report
+	readJSON(t, reportPath, &rpt)
+	audit := extractNormalizeAudit(t, rpt)
+	if audit.InputSegmentCount != 5 || audit.OutputSegmentCount != 3 {
+		t.Fatalf("audit counts = in:%d out:%d, want in:5 out:3", audit.InputSegmentCount, audit.OutputSegmentCount)
+	}
+	if audit.TimingFieldsRepaired != 2 {
+		t.Fatalf("timing fields repaired = %d, want 2", audit.TimingFieldsRepaired)
+	}
+	if audit.TimingOrderSwapped != 1 {
+		t.Fatalf("timing order swapped = %d, want 1", audit.TimingOrderSwapped)
+	}
+	if audit.SpeakerFilled != 2 {
+		t.Fatalf("speaker filled = %d, want 2", audit.SpeakerFilled)
+	}
+	if audit.SegmentsDroppedText != 2 {
+		t.Fatalf("segments dropped text = %d, want 2", audit.SegmentsDroppedText)
+	}
+}
+
 func TestNormalizeSelectedOutputSchemaIsHonored(t *testing.T) {
 	dir := t.TempDir()
 	input := writeJSONFile(t, dir, "input.json", `{"segments":[{"start":1,"end":2,"speaker":"A","text":"one"}]}`)
@@ -434,11 +494,16 @@ type normalizeAudit struct {
 	OutputFile             string   `json:"output_file"`
 	InputShape             string   `json:"input_shape"`
 	InputSegmentCount      int      `json:"input_segment_count"`
+	OutputSegmentCount     int      `json:"output_segment_count"`
 	OutputSchema           string   `json:"output_schema"`
 	OutputModules          []string `json:"output_modules"`
 	IDsReassigned          bool     `json:"ids_reassigned"`
 	SortingChangedInput    bool     `json:"sorting_changed_input_order"`
 	SegmentsWithCategories int      `json:"segments_with_categories"`
+	TimingFieldsRepaired   int      `json:"timing_fields_repaired"`
+	TimingOrderSwapped     int      `json:"timing_order_swapped"`
+	SpeakerFilled          int      `json:"speaker_filled"`
+	SegmentsDroppedText    int      `json:"segments_dropped_text"`
 }
 
 func extractNormalizeAudit(t *testing.T, rpt report.Report) normalizeAudit {

internal/normalize/build.go

@@ -15,6 +15,7 @@ import (
 // BuildResult contains normalize output plus deterministic transformation diagnostics.
 type BuildResult struct {
 	Output                 any
+	OutputSegmentCount     int
 	SortingChanged         bool
 	IDsReassigned          bool
 	SegmentsWithCategories int
@@ -35,6 +36,7 @@ func Build(parsed ParsedTranscript, cfg config.NormalizeConfig) (BuildResult, er
 		}
 		return BuildResult{
 			Output:                 output,
+			OutputSegmentCount:     len(ordered),
 			SortingChanged:         sortingChanged,
 			IDsReassigned:          idsReassigned,
 			SegmentsWithCategories: segmentsWithCategories,
@@ -46,6 +48,7 @@ func Build(parsed ParsedTranscript, cfg config.NormalizeConfig) (BuildResult, er
 		}
 		return BuildResult{
 			Output:                 output,
+			OutputSegmentCount:     len(ordered),
 			SortingChanged:         sortingChanged,
 			IDsReassigned:          idsReassigned,
 			SegmentsWithCategories: segmentsWithCategories,
@@ -57,6 +60,7 @@ func Build(parsed ParsedTranscript, cfg config.NormalizeConfig) (BuildResult, er
 		}
 		return BuildResult{
 			Output:                 output,
+			OutputSegmentCount:     len(ordered),
 			SortingChanged:         sortingChanged,
 			IDsReassigned:          idsReassigned,
 			SegmentsWithCategories: segmentsWithCategories,

internal/normalize/normalize.go

@@ -19,11 +19,16 @@ type normalizeAudit struct {
 	OutputFile             string   `json:"output_file"`
 	InputShape             string   `json:"input_shape"`
 	InputSegmentCount      int      `json:"input_segment_count"`
+	OutputSegmentCount     int      `json:"output_segment_count"`
 	OutputSchema           string   `json:"output_schema"`
 	OutputModules          []string `json:"output_modules"`
 	IDsReassigned          bool     `json:"ids_reassigned"`
 	SortingChangedInput    bool     `json:"sorting_changed_input_order"`
 	SegmentsWithCategories int      `json:"segments_with_categories"`
+	TimingFieldsRepaired   int      `json:"timing_fields_repaired"`
+	TimingOrderSwapped     int      `json:"timing_order_swapped"`
+	SpeakerFilled          int      `json:"speaker_filled"`
+	SegmentsDroppedText    int      `json:"segments_dropped_text"`
 }
 
 // Run executes artifact-level normalization.
@@ -52,12 +57,17 @@ func Run(ctx context.Context, cfg config.NormalizeConfig) error {
 			InputFile:              cfg.InputFile,
 			OutputFile:             cfg.OutputFile,
 			InputShape:             string(parsed.Shape),
-			InputSegmentCount:      len(parsed.Segments),
+			InputSegmentCount:      parsed.InputSegmentCount,
+			OutputSegmentCount:     built.OutputSegmentCount,
 			OutputSchema:           cfg.OutputSchema,
 			OutputModules:          append([]string(nil), cfg.OutputModules...),
 			IDsReassigned:          built.IDsReassigned,
 			SortingChangedInput:    built.SortingChanged,
 			SegmentsWithCategories: built.SegmentsWithCategories,
+			TimingFieldsRepaired:   parsed.Stats.TimingFieldsRepaired,
+			TimingOrderSwapped:     parsed.Stats.TimingOrderSwapped,
+			SpeakerFilled:          parsed.Stats.SpeakerFilled,
+			SegmentsDroppedText:    parsed.Stats.SegmentsDroppedText,
 		}
 		auditJSON, err := json.Marshal(audit)
 		if err != nil {
@@ -68,20 +78,24 @@ func Run(ctx context.Context, cfg config.NormalizeConfig) error {
 			report.Info("normalize", "normalize", "started normalize command"),
 			report.Info("normalize", "normalize", fmt.Sprintf("input file: %s", cfg.InputFile)),
 			report.Info("normalize", "normalize", fmt.Sprintf("detected input shape: %s", parsed.Shape)),
-			report.Info("normalize", "normalize", fmt.Sprintf("input segment count: %d", len(parsed.Segments))),
+			report.Info("normalize", "normalize", fmt.Sprintf("input segment count: %d", parsed.InputSegmentCount)),
 			report.Info("normalize", "normalize", fmt.Sprintf("selected output schema: %s", cfg.OutputSchema)),
 			report.Info("normalize", "normalize", fmt.Sprintf("selected output modules: %s", strings.Join(cfg.OutputModules, ","))),
 			report.Info("normalize", "normalize", fmt.Sprintf("output file: %s", cfg.OutputFile)),
 			report.Info("normalize", "normalize", fmt.Sprintf("ids reassigned: %t", built.IDsReassigned)),
 			report.Info("normalize", "normalize", fmt.Sprintf("sorting changed input order: %t", built.SortingChanged)),
 			report.Info("normalize", "normalize", fmt.Sprintf("segments with categories: %d", built.SegmentsWithCategories)),
+			report.Info("normalize", "normalize", fmt.Sprintf("timing fields repaired: %d", parsed.Stats.TimingFieldsRepaired)),
+			report.Info("normalize", "normalize", fmt.Sprintf("timing order swapped: %d", parsed.Stats.TimingOrderSwapped)),
+			report.Info("normalize", "normalize", fmt.Sprintf("speaker placeholders added: %d", parsed.Stats.SpeakerFilled)),
+			report.Info("normalize", "normalize", fmt.Sprintf("segments dropped for empty text: %d", parsed.Stats.SegmentsDroppedText)),
 			report.Info("normalize", "normalize-audit", string(auditJSON)),
 		}
-		if len(parsed.Segments) == 0 {
+		if parsed.InputSegmentCount == 0 {
 			events = append(events, report.Warning("normalize", "normalize", "input transcript contains zero segments"))
 		}
 		events = append(events,
-			report.Info("normalize", "validate-output", fmt.Sprintf("validated %d output segment(s)", len(parsed.Segments))),
+			report.Info("normalize", "validate-output", fmt.Sprintf("validated %d output segment(s)", built.OutputSegmentCount)),
 			report.Info("output", "json", "wrote transcript JSON"),
 		)
 

internal/normalize/parse.go

@@ -19,14 +19,20 @@ const (
 
 // ParsedTranscript is the validated normalize input model.
 type ParsedTranscript struct {
-	Shape    InputShape
-	Segments []InputSegment
+	Shape             InputShape
+	InputSegmentCount int
+	Stats             NormalizeStats
+	Segments          []InputSegment
 }
 
 // InputSegment is a validated segment from normalize input.
 type InputSegment struct {
 	InputIndex         int
 	OriginalID         *int
+	StartPresent       bool
+	EndPresent         bool
+	SpeakerPresent     bool
+	TextPresent        bool
 	Start              float64
 	End                float64
 	Speaker            string
@@ -39,6 +45,14 @@ type InputSegment struct {
 	OverlapGroupID     *int
 }
 
+// NormalizeStats captures deterministic repair/drop outcomes from normalize input processing.
+type NormalizeStats struct {
+	TimingFieldsRepaired int
+	TimingOrderSwapped   int
+	SpeakerFilled        int
+	SegmentsDroppedText  int
+}
+
 type inputSegmentPayload struct {
 	ID                 *int     `json:"id"`
 	Start              *float64 `json:"start"`
@@ -85,13 +99,15 @@ func ParseReader(reader io.Reader) (ParsedTranscript, error) {
 	case '{':
 		return parseObjectShape(trimmed)
 	case '[':
-		segments, err := parseSegmentsArray(trimmed)
+		segments, stats, inputCount, err := parseSegmentsArray(trimmed)
 		if err != nil {
 			return ParsedTranscript{}, err
 		}
 		return ParsedTranscript{
-			Shape:    ShapeBareSegmentsArray,
-			Segments: segments,
+			Shape:             ShapeBareSegmentsArray,
+			InputSegmentCount: inputCount,
+			Stats:             stats,
+			Segments:          segments,
 		}, nil
 	default:
 		return ParsedTranscript{}, fmt.Errorf("normalize input must be a top-level object with \"segments\" or a top-level segment array")
@@ -121,72 +137,73 @@ func parseObjectShape(raw []byte) (ParsedTranscript, error) {
 		return ParsedTranscript{}, fmt.Errorf("normalize object input must contain a \"segments\" field")
 	}
 
-	segments, err := parseSegmentsArray(segmentsRaw)
+	segments, stats, inputCount, err := parseSegmentsArray(segmentsRaw)
 	if err != nil {
 		return ParsedTranscript{}, err
 	}
 
 	return ParsedTranscript{
-		Shape:    ShapeObjectWithSegments,
-		Segments: segments,
+		Shape:             ShapeObjectWithSegments,
+		InputSegmentCount: inputCount,
+		Stats:             stats,
+		Segments:          segments,
 	}, nil
 }
 
-func parseSegmentsArray(raw []byte) ([]InputSegment, error) {
+func parseSegmentsArray(raw []byte) ([]InputSegment, NormalizeStats, int, error) {
 	var segmentValues []json.RawMessage
 	if err := json.Unmarshal(raw, &segmentValues); err != nil {
-		return nil, fmt.Errorf("normalize input \"segments\" must be an array")
+		return nil, NormalizeStats{}, 0, fmt.Errorf("normalize input \"segments\" must be an array")
 	}
 
 	segments := make([]InputSegment, len(segmentValues))
 	for index, segmentRaw := range segmentValues {
-		segment, err := parseSegment(index, segmentRaw)
+		segment, err := decodeSegment(index, segmentRaw)
 		if err != nil {
-			return nil, err
+			return nil, NormalizeStats{}, 0, err
 		}
 		segments[index] = segment
 	}
-	return segments, nil
+	normalized, stats, err := normalizeSegments(segments)
+	return normalized, stats, len(segmentValues), err
 }
 
-func parseSegment(index int, raw []byte) (InputSegment, error) {
+func decodeSegment(index int, raw []byte) (InputSegment, error) {
 	var payload inputSegmentPayload
 	if err := json.Unmarshal(raw, &payload); err != nil {
 		return InputSegment{}, fmt.Errorf("segment %d: invalid segment object: %w", index, err)
 	}
 
-	if payload.Start == nil {
-		return InputSegment{}, fmt.Errorf("segment %d is missing required field \"start\"", index)
+	var start float64
+	if payload.Start != nil {
+		start = *payload.Start
 	}
-	if payload.End == nil {
-		return InputSegment{}, fmt.Errorf("segment %d is missing required field \"end\"", index)
-	}
-	if payload.Speaker == nil {
-		return InputSegment{}, fmt.Errorf("segment %d is missing required field \"speaker\"", index)
-	}
-	if payload.Text == nil {
-		return InputSegment{}, fmt.Errorf("segment %d is missing required field \"text\"", index)
+	var end float64
+	if payload.End != nil {
+		end = *payload.End
 	}
 
-	if *payload.Start < 0 {
-		return InputSegment{}, fmt.Errorf("segment %d has start %v; start must be >= 0", index, *payload.Start)
-	}
-	if *payload.End < *payload.Start {
-		return InputSegment{}, fmt.Errorf("segment %d has end %v before start %v", index, *payload.End, *payload.Start)
+	speaker := ""
+	if payload.Speaker != nil {
+		speaker = strings.TrimSpace(*payload.Speaker)
 	}
 
-	speaker := strings.TrimSpace(*payload.Speaker)
-	if speaker == "" {
-		return InputSegment{}, fmt.Errorf("segment %d has empty \"speaker\"; speaker must be non-empty", index)
+	text := ""
+	if payload.Text != nil {
+		text = *payload.Text
 	}
 
 	return InputSegment{
 		InputIndex:         index,
 		OriginalID:         payload.ID,
-		Start:              *payload.Start,
-		End:                *payload.End,
+		StartPresent:       payload.Start != nil,
+		EndPresent:         payload.End != nil,
+		SpeakerPresent:     payload.Speaker != nil,
+		TextPresent:        payload.Text != nil,
+		Start:              start,
+		End:                end,
 		Speaker:            speaker,
-		Text:               *payload.Text,
+		Text:               text,
 		Categories:         append([]string(nil), payload.Categories...),
 		Source:             payload.Source,
 		SourceSegmentIndex: payload.SourceSegmentIndex,
@@ -195,3 +212,96 @@ func parseSegment(index int, raw []byte) (InputSegment, error) {
 		OverlapGroupID:     payload.OverlapGroupID,
 	}, nil
 }
+
+func normalizeSegments(segments []InputSegment) ([]InputSegment, NormalizeStats, error) {
+	stats := NormalizeStats{}
+	filtered := make([]InputSegment, 0, len(segments))
+	for _, segment := range segments {
+		if strings.TrimSpace(segment.Text) == "" {
+			stats.SegmentsDroppedText++
+			continue
+		}
+		filtered = append(filtered, segment)
+	}
+
+	if len(filtered) == 0 {
+		return filtered, stats, nil
+	}
+
+	hasStart := make([]bool, len(filtered))
+	hasEnd := make([]bool, len(filtered))
+	for index, segment := range filtered {
+		hasStart[index] = segment.StartPresent
+		hasEnd[index] = segment.EndPresent
+	}
+	for index := range filtered {
+		if !hasStart[index] && !hasEnd[index] {
+			midpoint := inferTimestampFromNeighbors(filtered, hasStart, hasEnd, index)
+			filtered[index].Start = midpoint
+			filtered[index].End = midpoint
+			stats.TimingFieldsRepaired += 2
+			hasStart[index] = true
+			hasEnd[index] = true
+			continue
+		}
+		if hasStart[index] && !hasEnd[index] {
+			filtered[index].End = filtered[index].Start
+			stats.TimingFieldsRepaired++
+			hasEnd[index] = true
+		}
+		if !hasStart[index] && hasEnd[index] {
+			filtered[index].Start = filtered[index].End
+			stats.TimingFieldsRepaired++
+			hasStart[index] = true
+		}
+	}
+
+	for index := range filtered {
+		if filtered[index].Start < 0 {
+			return nil, NormalizeStats{}, fmt.Errorf("segment %d has start %v; start must be >= 0", filtered[index].InputIndex, filtered[index].Start)
+		}
+		if filtered[index].End < filtered[index].Start {
+			filtered[index].Start, filtered[index].End = filtered[index].End, filtered[index].Start
+			stats.TimingOrderSwapped++
+		}
+		if strings.TrimSpace(filtered[index].Speaker) == "" {
+			filtered[index].Speaker = "Unknown_Speaker"
+			stats.SpeakerFilled++
+		}
+	}
+
+	return filtered, stats, nil
+}
+
+func inferTimestampFromNeighbors(segments []InputSegment, hasStart []bool, hasEnd []bool, index int) float64 {
+	prevEnd, hasPrev := nearestPreviousEnd(segments, hasEnd, index)
+	nextStart, hasNext := nearestNextStart(segments, hasStart, index)
+	if hasPrev && hasNext {
+		return (prevEnd + nextStart) / 2
+	}
+	if hasPrev {
+		return prevEnd
+	}
+	if hasNext {
+		return nextStart
+	}
+	return 0
+}
+
+func nearestPreviousEnd(segments []InputSegment, hasEnd []bool, index int) (float64, bool) {
+	for i := index - 1; i >= 0; i-- {
+		if hasEnd[i] {
+			return segments[i].End, true
+		}
+	}
+	return 0, false
+}
+
+func nearestNextStart(segments []InputSegment, hasStart []bool, index int) (float64, bool) {
+	for i := index + 1; i < len(segments); i++ {
+		if hasStart[i] {
+			return segments[i].Start, true
+		}
+	}
+	return 0, false
+}

internal/normalize/parse_test.go

@@ -92,34 +92,70 @@ func TestParseReaderTopLevelScalarShapesFail(t *testing.T) {
 	}
 }
 
-func TestParseReaderMissingStartFails(t *testing.T) {
-	_, err := ParseReader(strings.NewReader(`[{"end":2,"speaker":"A","text":"t"}]`))
-	assertContains(t, err, `missing required field "start"`)
+func TestParseReaderMissingStartUsesEndValue(t *testing.T) {
+	parsed, err := ParseReader(strings.NewReader(`[{"end":2,"speaker":"A","text":"t"}]`))
+	if err != nil {
+		t.Fatalf("parse failed: %v", err)
+	}
+	if len(parsed.Segments) != 1 {
+		t.Fatalf("segment count = %d, want 1", len(parsed.Segments))
+	}
+	if parsed.Segments[0].Start != 2 || parsed.Segments[0].End != 2 {
+		t.Fatalf("segment timing = %v..%v, want 2..2", parsed.Segments[0].Start, parsed.Segments[0].End)
+	}
 }
 
-func TestParseReaderMissingEndFails(t *testing.T) {
-	_, err := ParseReader(strings.NewReader(`[{"start":1,"speaker":"A","text":"t"}]`))
-	assertContains(t, err, `missing required field "end"`)
+func TestParseReaderMissingEndUsesStartValue(t *testing.T) {
+	parsed, err := ParseReader(strings.NewReader(`[{"start":1,"speaker":"A","text":"t"}]`))
+	if err != nil {
+		t.Fatalf("parse failed: %v", err)
+	}
+	if len(parsed.Segments) != 1 {
+		t.Fatalf("segment count = %d, want 1", len(parsed.Segments))
+	}
+	if parsed.Segments[0].Start != 1 || parsed.Segments[0].End != 1 {
+		t.Fatalf("segment timing = %v..%v, want 1..1", parsed.Segments[0].Start, parsed.Segments[0].End)
+	}
 }
 
-func TestParseReaderMissingSpeakerFails(t *testing.T) {
-	_, err := ParseReader(strings.NewReader(`[{"start":1,"end":2,"text":"t"}]`))
-	assertContains(t, err, `missing required field "speaker"`)
+func TestParseReaderMissingSpeakerUsesUnknownPlaceholder(t *testing.T) {
+	parsed, err := ParseReader(strings.NewReader(`[{"start":1,"end":2,"text":"t"}]`))
+	if err != nil {
+		t.Fatalf("parse failed: %v", err)
+	}
+	if parsed.Segments[0].Speaker != "Unknown_Speaker" {
+		t.Fatalf("speaker = %q, want Unknown_Speaker", parsed.Segments[0].Speaker)
+	}
 }
 
-func TestParseReaderEmptySpeakerFails(t *testing.T) {
-	_, err := ParseReader(strings.NewReader(`[{"start":1,"end":2,"speaker":"   ","text":"t"}]`))
-	assertContains(t, err, `speaker must be non-empty`)
+func TestParseReaderEmptySpeakerUsesUnknownPlaceholder(t *testing.T) {
+	parsed, err := ParseReader(strings.NewReader(`[{"start":1,"end":2,"speaker":"   ","text":"t"}]`))
+	if err != nil {
+		t.Fatalf("parse failed: %v", err)
+	}
+	if parsed.Segments[0].Speaker != "Unknown_Speaker" {
+		t.Fatalf("speaker = %q, want Unknown_Speaker", parsed.Segments[0].Speaker)
+	}
 }
 
-func TestParseReaderMissingTextFails(t *testing.T) {
-	_, err := ParseReader(strings.NewReader(`[{"start":1,"end":2,"speaker":"A"}]`))
-	assertContains(t, err, `missing required field "text"`)
+func TestParseReaderMissingTextDropsSegment(t *testing.T) {
+	parsed, err := ParseReader(strings.NewReader(`[{"start":1,"end":2,"speaker":"A"}]`))
+	if err != nil {
+		t.Fatalf("parse failed: %v", err)
+	}
+	if len(parsed.Segments) != 0 {
+		t.Fatalf("segment count = %d, want 0", len(parsed.Segments))
+	}
 }
 
-func TestParseReaderEndBeforeStartFails(t *testing.T) {
-	_, err := ParseReader(strings.NewReader(`[{"start":3,"end":2,"speaker":"A","text":"t"}]`))
-	assertContains(t, err, "before start")
+func TestParseReaderEndBeforeStartSwapsValues(t *testing.T) {
+	parsed, err := ParseReader(strings.NewReader(`[{"start":3,"end":2,"speaker":"A","text":"t"}]`))
+	if err != nil {
+		t.Fatalf("parse failed: %v", err)
+	}
+	if parsed.Segments[0].Start != 2 || parsed.Segments[0].End != 3 {
+		t.Fatalf("segment timing = %v..%v, want 2..3", parsed.Segments[0].Start, parsed.Segments[0].End)
+	}
 }
 
 func TestParseReaderNegativeStartFails(t *testing.T) {
@@ -153,6 +189,68 @@ func TestParseReaderCategoriesPreservedWhenValid(t *testing.T) {
 	}
 }
 
+func TestParseReaderMissingBothTimesMiddleUsesNeighborMidpoint(t *testing.T) {
+	parsed, err := ParseReader(strings.NewReader(`[
+		{"start":1,"end":2,"speaker":"A","text":"left"},
+		{"speaker":"B","text":"middle"},
+		{"start":6,"end":7,"speaker":"C","text":"right"}
+	]`))
+	if err != nil {
+		t.Fatalf("parse failed: %v", err)
+	}
+	if parsed.Segments[1].Start != 4 || parsed.Segments[1].End != 4 {
+		t.Fatalf("middle timing = %v..%v, want 4..4", parsed.Segments[1].Start, parsed.Segments[1].End)
+	}
+}
+
+func TestParseReaderMissingBothTimesEdgeFallbacks(t *testing.T) {
+	parsedFirst, err := ParseReader(strings.NewReader(`[
+		{"speaker":"A","text":"first"},
+		{"start":5,"end":6,"speaker":"B","text":"second"}
+	]`))
+	if err != nil {
+		t.Fatalf("parse first failed: %v", err)
+	}
+	if parsedFirst.Segments[0].Start != 5 || parsedFirst.Segments[0].End != 5 {
+		t.Fatalf("first timing = %v..%v, want 5..5", parsedFirst.Segments[0].Start, parsedFirst.Segments[0].End)
+	}
+
+	parsedLast, err := ParseReader(strings.NewReader(`[
+		{"start":1,"end":2,"speaker":"A","text":"first"},
+		{"speaker":"B","text":"last"}
+	]`))
+	if err != nil {
+		t.Fatalf("parse last failed: %v", err)
+	}
+	if parsedLast.Segments[1].Start != 2 || parsedLast.Segments[1].End != 2 {
+		t.Fatalf("last timing = %v..%v, want 2..2", parsedLast.Segments[1].Start, parsedLast.Segments[1].End)
+	}
+}
+
+func TestParseReaderMissingBothTimesSingleSegmentUsesZero(t *testing.T) {
+	parsed, err := ParseReader(strings.NewReader(`[{"speaker":"A","text":"only"}]`))
+	if err != nil {
+		t.Fatalf("parse failed: %v", err)
+	}
+	if parsed.Segments[0].Start != 0 || parsed.Segments[0].End != 0 {
+		t.Fatalf("timing = %v..%v, want 0..0", parsed.Segments[0].Start, parsed.Segments[0].End)
+	}
+}
+
+func TestParseReaderEmptyOrWhitespaceTextDropped(t *testing.T) {
+	parsed, err := ParseReader(strings.NewReader(`[
+		{"start":1,"end":2,"speaker":"A","text":"ok"},
+		{"start":2,"end":3,"speaker":"A","text":""},
+		{"start":3,"end":4,"speaker":"A","text":"   "}
+	]`))
+	if err != nil {
+		t.Fatalf("parse failed: %v", err)
+	}
+	if len(parsed.Segments) != 1 {
+		t.Fatalf("segment count = %d, want 1", len(parsed.Segments))
+	}
+}
+
 func TestParseReaderOriginalInputIndexPreserved(t *testing.T) {
 	input := `[
 		{"start":1,"end":2,"speaker":"A","text":"one"},