seriatim/internal/cli/merge_test.go

package cli

import (
	"encoding/json"
	"os"
	"path/filepath"
	"strings"
	"testing"

	"gitea.maximumdirect.net/eric/seriatim/internal/model"
	"gitea.maximumdirect.net/eric/seriatim/internal/report"
)

func TestMergeWritesMergedOutputAndReport(t *testing.T) {
	dir := t.TempDir()
	inputA := writeJSONFile(t, dir, "a.json", `{
		"segments": [
			{"start": 10, "end": 11, "text": " second a ", "words": [{"word": "ignored"}]},
			{"start": 1, "end": 2, "text": "first a"}
		]
	}`)
	inputB := writeJSONFile(t, dir, "b.json", `{
		"segments": [
			{"start": 5, "end": 6, "text": "first b"}
		]
	}`)
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["a.json"]
  - speaker: Bob
    match: ["b.json"]
`)
	output := filepath.Join(dir, "merged.json")
	reportPath := filepath.Join(dir, "report.json")

	err := executeMerge(
		"--input-file", inputB,
		"--input-file", inputA,
		"--speakers", speakers,
		"--output-file", output,
		"--report-file", reportPath,
	)
	if err != nil {
		t.Fatalf("merge failed: %v", err)
	}

	var transcript model.FinalTranscript
	readJSON(t, output, &transcript)
	outputBytes, err := os.ReadFile(output)
	if err != nil {
		t.Fatalf("read output bytes: %v", err)
	}
	outputJSON := string(outputBytes)
	if !strings.Contains(outputJSON, `"overlap_groups": []`) {
		t.Fatalf("expected overlap_groups to serialize as an empty array, got:\n%s", outputJSON)
	}
	if transcript.Metadata.Application != "seriatim" {
		t.Fatalf("unexpected application metadata: %q", transcript.Metadata.Application)
	}
	if got, want := transcript.Metadata.InputFiles, []string{inputA, inputB}; !equalStrings(got, want) {
		t.Fatalf("input files not sorted deterministically: got %v want %v", got, want)
	}
	if got, want := len(transcript.Segments), 3; got != want {
		t.Fatalf("expected merged output to contain %d segments, got %d", want, got)
	}
	assertSegment(t, transcript.Segments[0], 1, inputA, 1, "Alice", 1, 2, "first a")
	assertSegment(t, transcript.Segments[1], 2, inputB, 0, "Bob", 5, 6, "first b")
	assertSegment(t, transcript.Segments[2], 3, inputA, 0, "Alice", 10, 11, "second a")
	if strings.Contains(outputJSON, "internal_ref") {
		t.Fatalf("did not expect internal_ref in output:\n%s", outputJSON)
	}
	if strings.Contains(outputJSON, "words") {
		t.Fatalf("did not expect words in output:\n%s", outputJSON)
	}
	if len(transcript.OverlapGroups) != 0 {
		t.Fatalf("expected placeholder output to contain no overlap groups, got %d", len(transcript.OverlapGroups))
	}

	var rpt report.Report
	readJSON(t, reportPath, &rpt)
	gotModules := make([]string, 0, len(rpt.Events))
	for _, event := range rpt.Events {
		gotModules = append(gotModules, event.Module)
	}
	wantModules := []string{
		"json-files",
		"validate-raw",
		"normalize-speakers",
		"trim-text",
		"placeholder-merger",
		"detect-overlaps",
		"resolve-overlaps",
		"autocorrect",
		"assign-ids",
		"validate-output",
		"json",
	}
	if !equalStrings(gotModules, wantModules) {
		t.Fatalf("report event order mismatch:\ngot  %v\nwant %v", gotModules, wantModules)
	}
}

func TestMergeTieBreakOrder(t *testing.T) {
	dir := t.TempDir()
	inputA := writeJSONFile(t, dir, "a.json", `{
		"segments": [
			{"start": 1, "end": 4, "text": "a-late-end"},
			{"start": 1, "end": 2, "text": "a-index-one"}
		]
	}`)
	inputB := writeJSONFile(t, dir, "b.json", `{
		"segments": [
			{"start": 1, "end": 2, "text": "b-same-time"}
		]
	}`)
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["a.json"]
  - speaker: Bob
    match: ["b.json"]
`)
	output := filepath.Join(dir, "merged.json")

	err := executeMerge(
		"--input-file", inputB,
		"--input-file", inputA,
		"--speakers", speakers,
		"--output-file", output,
	)
	if err != nil {
		t.Fatalf("merge failed: %v", err)
	}

	var transcript model.FinalTranscript
	readJSON(t, output, &transcript)
	got := []string{
		transcript.Segments[0].Text,
		transcript.Segments[1].Text,
		transcript.Segments[2].Text,
	}
	want := []string{"a-index-one", "b-same-time", "a-late-end"}
	if !equalStrings(got, want) {
		t.Fatalf("tie-break order mismatch: got %v want %v", got, want)
	}
	for index, segment := range transcript.Segments {
		if segment.ID != index+1 {
			t.Fatalf("segment %d has id %d; want %d", index, segment.ID, index+1)
		}
	}
}

func TestSpeakerMatchingUsesFirstMatchingRuleCaseInsensitive(t *testing.T) {
	dir := t.TempDir()
	input := writeJSONFile(t, dir, "2026-04-19-Adam_Rakestraw.json", `{
		"segments": [
			{"start": 1, "end": 2, "text": "hello"}
		]
	}`)
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: First Match
    match: ["adam"]
  - speaker: Later Match
    match: ["Adam_Rakestraw"]
`)
	output := filepath.Join(dir, "merged.json")

	err := executeMerge(
		"--input-file", input,
		"--speakers", speakers,
		"--output-file", output,
	)
	if err != nil {
		t.Fatalf("merge failed: %v", err)
	}

	var transcript model.FinalTranscript
	readJSON(t, output, &transcript)
	if got, want := transcript.Segments[0].Speaker, "First Match"; got != want {
		t.Fatalf("speaker = %q, want %q", got, want)
	}
}

func TestUnknownModulesFailDuringValidation(t *testing.T) {
	dir := t.TempDir()
	input := writeJSONFile(t, dir, "input.json", `{"segments":[]}`)
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["input.json"]
`)
	output := filepath.Join(dir, "merged.json")

	tests := []struct {
		name string
		args []string
		want string
	}{
		{
			name: "input reader",
			args: []string{"--input-reader", "missing-reader"},
			want: `unknown input reader "missing-reader"`,
		},
		{
			name: "preprocessing",
			args: []string{"--preprocessing-modules", "validate-raw,missing-module"},
			want: `unknown preprocessing module "missing-module"`,
		},
		{
			name: "postprocessing",
			args: []string{"--postprocessing-modules", "missing-module"},
			want: `unknown postprocessing module "missing-module"`,
		},
		{
			name: "output",
			args: []string{"--output-modules", "missing-module"},
			want: `unknown output module "missing-module"`,
		},
	}

	for _, test := range tests {
		t.Run(test.name, func(t *testing.T) {
			args := []string{
				"--input-file", input,
				"--speakers", speakers,
				"--output-file", output,
			}
			args = append(args, test.args...)

			err := executeMerge(args...)
			if err == nil {
				t.Fatal("expected error")
			}
			if !strings.Contains(err.Error(), test.want) {
				t.Fatalf("expected error to contain %q, got %q", test.want, err.Error())
			}
		})
	}
}

func TestInvalidPreprocessingOrderFails(t *testing.T) {
	dir := t.TempDir()
	input := writeJSONFile(t, dir, "input.json", `{"segments":[]}`)
	output := filepath.Join(dir, "merged.json")

	err := executeMerge(
		"--input-file", input,
		"--output-file", output,
		"--preprocessing-modules", "trim-text,validate-raw",
	)
	if err == nil {
		t.Fatal("expected error")
	}
	if !strings.Contains(err.Error(), `requires state "canonical"`) {
		t.Fatalf("unexpected error: %v", err)
	}
}

func TestMissingInputFileFailsBeforePipelineExecution(t *testing.T) {
	dir := t.TempDir()
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["missing.json"]
`)
	output := filepath.Join(dir, "merged.json")

	err := executeMerge(
		"--input-file", filepath.Join(dir, "missing.json"),
		"--speakers", speakers,
		"--output-file", output,
	)
	if err == nil {
		t.Fatal("expected error")
	}
	if !strings.Contains(err.Error(), "--input-file") {
		t.Fatalf("unexpected error: %v", err)
	}
}

func TestDefaultMergeWorksWithoutSpeakersOrAutocorrect(t *testing.T) {
	dir := t.TempDir()
	input := writeJSONFile(t, dir, "input.json", `{"segments":[{"start":1,"end":2,"text":"Frank"}]}`)
	output := filepath.Join(dir, "merged.json")
	reportPath := filepath.Join(dir, "report.json")

	err := executeMerge(
		"--input-file", input,
		"--output-file", output,
		"--report-file", reportPath,
	)
	if err != nil {
		t.Fatalf("merge failed: %v", err)
	}

	var transcript model.FinalTranscript
	readJSON(t, output, &transcript)
	if got, want := transcript.Segments[0].Speaker, "input.json"; got != want {
		t.Fatalf("speaker = %q, want %q", got, want)
	}
	if got, want := transcript.Segments[0].Text, "Frank"; got != want {
		t.Fatalf("text = %q, want %q", got, want)
	}

	var rpt report.Report
	readJSON(t, reportPath, &rpt)
	if !hasReportEvent(rpt, "preprocessing", "normalize-speakers", "using input basenames") {
		t.Fatal("expected normalize-speakers fallback report event")
	}
	if !hasReportEvent(rpt, "postprocessing", "autocorrect", "skipped autocorrect") {
		t.Fatal("expected autocorrect skip report event")
	}
}

func TestPreprocessingAutocorrectIsUnknownModule(t *testing.T) {
	dir := t.TempDir()
	input := writeJSONFile(t, dir, "input.json", `{"segments":[]}`)
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["input.json"]
`)
	autocorrect := writeYAMLFile(t, dir, "autocorrect.yml", `autocorrect:
  - target: Hrank
    match: ["Frank"]
`)
	output := filepath.Join(dir, "merged.json")

	err := executeMerge(
		"--input-file", input,
		"--speakers", speakers,
		"--autocorrect", autocorrect,
		"--output-file", output,
		"--preprocessing-modules", "validate-raw,normalize-speakers,autocorrect",
	)
	if err == nil {
		t.Fatal("expected error")
	}
	if !strings.Contains(err.Error(), `unknown preprocessing module "autocorrect"`) {
		t.Fatalf("unexpected error: %v", err)
	}
}

func TestPostprocessingAutocorrectUpdatesOutputAndReport(t *testing.T) {
	dir := t.TempDir()
	input := writeJSONFile(t, dir, "input.json", `{
		"segments": [
			{"start": 1, "end": 2, "text": "Frank met Mike Pat, not Franklin."},
			{"start": 3, "end": 4, "text": "God-free and FRANK stayed."}
		]
	}`)
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["input.json"]
`)
	autocorrect := writeYAMLFile(t, dir, "autocorrect.yml", `autocorrect:
  - target: Hrank
    match: ["Frank"]
  - target: Mike Brown
    match: ["Mike Pat"]
  - target: Godfrey
    match: ["God-free"]
`)
	output := filepath.Join(dir, "merged.json")
	reportPath := filepath.Join(dir, "report.json")

	err := executeMerge(
		"--input-file", input,
		"--speakers", speakers,
		"--autocorrect", autocorrect,
		"--output-file", output,
		"--report-file", reportPath,
		"--postprocessing-modules", "detect-overlaps,resolve-overlaps,autocorrect,assign-ids,validate-output",
	)
	if err != nil {
		t.Fatalf("merge failed: %v", err)
	}

	var transcript model.FinalTranscript
	readJSON(t, output, &transcript)
	if got, want := transcript.Segments[0].Text, "Hrank met Mike Brown, not Franklin."; got != want {
		t.Fatalf("segment 0 text = %q, want %q", got, want)
	}
	if got, want := transcript.Segments[1].Text, "Godfrey and FRANK stayed."; got != want {
		t.Fatalf("segment 1 text = %q, want %q", got, want)
	}

	var rpt report.Report
	readJSON(t, reportPath, &rpt)
	found := false
	for _, event := range rpt.Events {
		if event.Stage == "postprocessing" && event.Module == "autocorrect" {
			found = true
			if !strings.Contains(event.Message, "applied 3 autocorrect replacement(s)") {
				t.Fatalf("unexpected autocorrect report message: %q", event.Message)
			}
		}
	}
	if !found {
		t.Fatal("expected autocorrect report event")
	}
}

func TestInvalidAutocorrectFileFailsWhenProvided(t *testing.T) {
	dir := t.TempDir()
	input := writeJSONFile(t, dir, "input.json", `{"segments":[{"start":1,"end":2,"text":"Frank"}]}`)
	output := filepath.Join(dir, "merged.json")
	autocorrect := writeYAMLFile(t, dir, "autocorrect.yml", `autocorrect:
  - target: ""
    match: ["Frank"]
`)

	err := executeMerge(
		"--input-file", input,
		"--autocorrect", autocorrect,
		"--output-file", output,
	)
	if err == nil {
		t.Fatal("expected error")
	}
	if !strings.Contains(err.Error(), "must include target") {
		t.Fatalf("unexpected error: %v", err)
	}
}

func TestOutputJSONIsByteStable(t *testing.T) {
	dir := t.TempDir()
	inputA := writeJSONFile(t, dir, "a.json", `{"segments":[{"start":2,"end":3,"text":"a"}]}`)
	inputB := writeJSONFile(t, dir, "b.json", `{"segments":[{"start":1,"end":2,"text":"b"}]}`)
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["a.json"]
  - speaker: Bob
    match: ["b.json"]
`)
	outputA := filepath.Join(dir, "merged-a.json")
	outputB := filepath.Join(dir, "merged-b.json")

	args := []string{
		"--input-file", inputB,
		"--input-file", inputA,
		"--speakers", speakers,
	}

	err := executeMerge(append(append([]string(nil), args...), "--output-file", outputA)...)
	if err != nil {
		t.Fatalf("first merge failed: %v", err)
	}
	err = executeMerge(append(append([]string(nil), args...), "--output-file", outputB)...)
	if err != nil {
		t.Fatalf("second merge failed: %v", err)
	}

	first, err := os.ReadFile(outputA)
	if err != nil {
		t.Fatalf("read first output: %v", err)
	}
	second, err := os.ReadFile(outputB)
	if err != nil {
		t.Fatalf("read second output: %v", err)
	}
	if string(first) != string(second) {
		t.Fatalf("expected byte-stable output\nfirst:\n%s\nsecond:\n%s", first, second)
	}
}

func TestMissingSpeakerMappingFails(t *testing.T) {
	dir := t.TempDir()
	input := writeJSONFile(t, dir, "input.json", `{"segments":[]}`)
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["other.json"]
`)
	output := filepath.Join(dir, "merged.json")

	err := executeMerge(
		"--input-file", input,
		"--speakers", speakers,
		"--output-file", output,
	)
	if err == nil {
		t.Fatal("expected error")
	}
	if !strings.Contains(err.Error(), `speaker map has no match for input basename "input.json"`) {
		t.Fatalf("unexpected error: %v", err)
	}
}

func TestMalformedJSONFails(t *testing.T) {
	dir := t.TempDir()
	input := writeJSONFile(t, dir, "input.json", `{"segments":[`)
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["input.json"]
`)
	output := filepath.Join(dir, "merged.json")

	err := executeMerge(
		"--input-file", input,
		"--speakers", speakers,
		"--output-file", output,
	)
	if err == nil {
		t.Fatal("expected error")
	}
	if !strings.Contains(err.Error(), "parse input file") {
		t.Fatalf("unexpected error: %v", err)
	}
}

func TestMissingTopLevelSegmentsFails(t *testing.T) {
	dir := t.TempDir()
	input := writeJSONFile(t, dir, "input.json", `{}`)
	speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["input.json"]
`)
	output := filepath.Join(dir, "merged.json")

	err := executeMerge(
		"--input-file", input,
		"--speakers", speakers,
		"--output-file", output,
	)
	if err == nil {
		t.Fatal("expected error")
	}
	if !strings.Contains(err.Error(), "must contain top-level segments array") {
		t.Fatalf("unexpected error: %v", err)
	}
}

func TestInvalidSegmentFieldsFailWithSourceAndIndex(t *testing.T) {
	tests := []struct {
		name string
		json string
		want string
	}{
		{
			name: "missing start",
			json: `{"segments":[{"end":1,"text":"x"}]}`,
			want: "segment 0 missing numeric start",
		},
		{
			name: "wrong typed end",
			json: `{"segments":[{"start":0,"end":"1","text":"x"}]}`,
			want: "segment 0 end must be numeric",
		},
		{
			name: "wrong typed text",
			json: `{"segments":[{"start":0,"end":1,"text":7}]}`,
			want: "segment 0 text must be a string",
		},
		{
			name: "null text",
			json: `{"segments":[{"start":0,"end":1,"text":null}]}`,
			want: "segment 0 missing string text",
		},
	}

	for _, test := range tests {
		t.Run(test.name, func(t *testing.T) {
			dir := t.TempDir()
			input := writeJSONFile(t, dir, "input.json", test.json)
			speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["input.json"]
`)
			output := filepath.Join(dir, "merged.json")

			err := executeMerge(
				"--input-file", input,
				"--speakers", speakers,
				"--output-file", output,
			)
			if err == nil {
				t.Fatal("expected error")
			}
			if !strings.Contains(err.Error(), input) {
				t.Fatalf("expected error to contain source path %q, got %v", input, err)
			}
			if !strings.Contains(err.Error(), test.want) {
				t.Fatalf("expected error to contain %q, got %v", test.want, err)
			}
		})
	}
}

func TestInvalidTimingFails(t *testing.T) {
	tests := []struct {
		name string
		json string
		want string
	}{
		{
			name: "negative start",
			json: `{"segments":[{"start":-1,"end":1,"text":"x"}]}`,
			want: "segment 0 has negative start",
		},
		{
			name: "end before start",
			json: `{"segments":[{"start":2,"end":1,"text":"x"}]}`,
			want: "segment 0 has end before start",
		},
	}

	for _, test := range tests {
		t.Run(test.name, func(t *testing.T) {
			dir := t.TempDir()
			input := writeJSONFile(t, dir, "input.json", test.json)
			speakers := writeYAMLFile(t, dir, "speakers.yml", `match:
  - speaker: Alice
    match: ["input.json"]
`)
			output := filepath.Join(dir, "merged.json")

			err := executeMerge(
				"--input-file", input,
				"--speakers", speakers,
				"--output-file", output,
			)
			if err == nil {
				t.Fatal("expected error")
			}
			if !strings.Contains(err.Error(), test.want) {
				t.Fatalf("expected error to contain %q, got %v", test.want, err)
			}
		})
	}
}

func executeMerge(args ...string) error {
	cmd := NewRootCommand()
	cmd.SetArgs(append([]string{"merge"}, args...))
	return cmd.Execute()
}

func writeJSONFile(t *testing.T, dir string, name string, content string) string {
	t.Helper()

	path := filepath.Join(dir, name)
	if err := os.WriteFile(path, []byte(content+"\n"), 0o600); err != nil {
		t.Fatalf("write file: %v", err)
	}
	return path
}

func writeYAMLFile(t *testing.T, dir string, name string, content string) string {
	t.Helper()

	path := filepath.Join(dir, name)
	if err := os.WriteFile(path, []byte(content), 0o600); err != nil {
		t.Fatalf("write file: %v", err)
	}
	return path
}

func readJSON(t *testing.T, path string, target any) {
	t.Helper()

	data, err := os.ReadFile(path)
	if err != nil {
		t.Fatalf("read %s: %v", path, err)
	}
	if err := json.Unmarshal(data, target); err != nil {
		t.Fatalf("unmarshal %s: %v", path, err)
	}
}

func equalStrings(left []string, right []string) bool {
	if len(left) != len(right) {
		return false
	}
	for index := range left {
		if left[index] != right[index] {
			return false
		}
	}
	return true
}

func hasReportEvent(rpt report.Report, stage string, module string, messageSubstring string) bool {
	for _, event := range rpt.Events {
		if event.Stage == stage && event.Module == module && strings.Contains(event.Message, messageSubstring) {
			return true
		}
	}
	return false
}

func assertSegment(t *testing.T, segment model.Segment, id int, source string, sourceIndex int, speaker string, start float64, end float64, text string) {
	t.Helper()

	if segment.ID != id {
		t.Fatalf("segment ID = %d, want %d", segment.ID, id)
	}
	if segment.Source != source {
		t.Fatalf("segment source = %q, want %q", segment.Source, source)
	}
	if segment.SourceSegmentIndex != sourceIndex {
		t.Fatalf("segment source index = %d, want %d", segment.SourceSegmentIndex, sourceIndex)
	}
	if segment.Speaker != speaker {
		t.Fatalf("segment speaker = %q, want %q", segment.Speaker, speaker)
	}
	if segment.Start != start {
		t.Fatalf("segment start = %f, want %f", segment.Start, start)
	}
	if segment.End != end {
		t.Fatalf("segment end = %f, want %f", segment.End, end)
	}
	if segment.Text != text {
		t.Fatalf("segment text = %q, want %q", segment.Text, text)
	}
}