eric/seriatim
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 4 Wiki Activity

Added a module to coalesce adjacent same-speaker segments

Browse Source
This commit is contained in:
Eric Rakestraw
2026-04-27 19:30:00 -05:00
parent 13d972cb24
commit aab6d12730
12 changed files with 919 additions and 28 deletions
Download Patch File Download Diff File Expand all files Collapse all files

1
internal/cli/merge.go
View File

@@ -34,6 +34,7 @@ func newMergeCommand() *cobra.Command {
flags.StringVar(&opts.OutputModules, "output-modules", config.DefaultOutputModules, "comma-separated output modules")
flags.StringVar(&opts.PreprocessingModules, "preprocessing-modules", config.DefaultPreprocessingModules, "comma-separated preprocessing modules")
flags.StringVar(&opts.PostprocessingModules, "postprocessing-modules", config.DefaultPostprocessingModules, "comma-separated postprocessing modules")
flags.StringVar(&opts.CoalesceGap, "coalesce-gap", config.DefaultCoalesceGapValue, "maximum same-speaker gap in seconds for coalesce")
return cmd
}

177
internal/cli/merge_test.go
View File

@@ -90,6 +90,7 @@ func TestMergeWritesMergedOutputAndReport(t *testing.T) {
"placeholder-merger",
"detect-overlaps",
"resolve-overlaps",
"coalesce",
"detect-overlaps",
"autocorrect",
"assign-ids",
@@ -195,7 +196,7 @@ func TestMergeDetectsOverlapGroups(t *testing.T) {
if group.Start != 1 || group.End != 6 {
t.Fatalf("group bounds = %f-%f, want 1-6", group.Start, group.End)
}
wantRefs := []string{inputA + "#0", inputA + "#1", inputB + "#0"}
wantRefs := []string{"coalesce:1", inputB + "#0"}
if !equalStrings(group.Segments, wantRefs) {
t.Fatalf("group refs = %v, want %v", group.Segments, wantRefs)
}
@@ -210,6 +211,15 @@ func TestMergeDetectsOverlapGroups(t *testing.T) {
t.Fatalf("segment %d overlap group ID = %d, want 1", index, segment.OverlapGroupID)
}
}
if len(transcript.Segments) != 2 {
t.Fatalf("segment count = %d, want 2", len(transcript.Segments))
}
if transcript.Segments[0].SourceRef != "coalesce:1" {
t.Fatalf("coalesced source_ref = %q, want coalesce:1", transcript.Segments[0].SourceRef)
}
if !equalStrings(transcript.Segments[0].DerivedFrom, []string{inputA + "#0", inputA + "#1"}) {
t.Fatalf("coalesced derived_from = %v", transcript.Segments[0].DerivedFrom)
}
var rpt report.Report
readJSON(t, reportPath, &rpt)
@@ -410,6 +420,171 @@ func TestMergeDetectsResidualOverlapsAfterResolution(t *testing.T) {
}
}
func TestMergeOrdersNearStartWordRunsShorterFirst(t *testing.T) {
dir := t.TempDir()
inputA := writeJSONFile(t, dir, "a.json", `{
"segments": [
{
"start": 1,
"end": 4,
"text": "alice long",
"words": [
{"word": "alice-long", "start": 1.0, "end": 2.0}
]
}
]
}`)
inputB := writeJSONFile(t, dir, "b.json", `{
"segments": [
{
"start": 1.1,
"end": 3,
"text": "bob short",
"words": [
{"word": "bob-short", "start": 1.2, "end": 1.3}
]
}
]
}`)
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", inputA,
"--input-file", inputB,
"--speakers", speakers,
"--output-file", output,
)
if err != nil {
t.Fatalf("merge failed: %v", err)
}
var transcript model.FinalTranscript
readJSON(t, output, &transcript)
if len(transcript.Segments) != 2 {
t.Fatalf("segment count = %d, want 2", len(transcript.Segments))
}
if transcript.Segments[0].Text != "bob-short" || transcript.Segments[0].ID != 1 {
t.Fatalf("first segment = %#v, want bob-short with ID 1", transcript.Segments[0])
}
if transcript.Segments[1].Text != "alice-long" || transcript.Segments[1].ID != 2 {
t.Fatalf("second segment = %#v, want alice-long with ID 2", transcript.Segments[1])
}
if len(transcript.OverlapGroups) != 1 {
t.Fatalf("overlap group count = %d, want 1", len(transcript.OverlapGroups))
}
if transcript.Segments[0].OverlapGroupID != 1 || transcript.Segments[1].OverlapGroupID != 1 {
t.Fatalf("segments should retain residual overlap annotation: %#v", transcript.Segments)
}
wantRefs := []string{"word-run:1:1:1", "word-run:1:2:1"}
if !equalStrings(transcript.OverlapGroups[0].Segments, wantRefs) {
t.Fatalf("overlap refs = %v, want %v", transcript.OverlapGroups[0].Segments, wantRefs)
}
}
func TestMergeCoalescesSameSpeakerSegmentsBeforeFinalOverlapDetection(t *testing.T) {
dir := t.TempDir()
inputA := writeJSONFile(t, dir, "a.json", `{
"segments": [
{"start": 1, "end": 2, "text": "first"},
{"start": 4, "end": 5, "text": "second"}
]
}`)
inputB := writeJSONFile(t, dir, "b.json", `{
"segments": [
{"start": 4.5, "end": 6, "text": "bob"}
]
}`)
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", inputA,
"--input-file", inputB,
"--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)
if len(transcript.Segments) != 2 {
t.Fatalf("segment count = %d, want 2", len(transcript.Segments))
}
alice := transcript.Segments[0]
if alice.ID != 1 || alice.Text != "first second" || alice.SourceRef != "coalesce:1" {
t.Fatalf("unexpected coalesced Alice segment: %#v", alice)
}
if alice.SourceSegmentIndex != nil {
t.Fatalf("coalesced segment source_segment_index = %d, want nil", *alice.SourceSegmentIndex)
}
if !equalStrings(alice.DerivedFrom, []string{inputA + "#0", inputA + "#1"}) {
t.Fatalf("derived_from = %v", alice.DerivedFrom)
}
if transcript.Segments[1].ID != 2 || transcript.Segments[1].Text != "bob" {
t.Fatalf("unexpected Bob segment: %#v", transcript.Segments[1])
}
if len(transcript.OverlapGroups) != 1 {
t.Fatalf("overlap group count = %d, want 1", len(transcript.OverlapGroups))
}
group := transcript.OverlapGroups[0]
if !equalStrings(group.Segments, []string{"coalesce:1", inputB + "#0"}) {
t.Fatalf("group refs = %v", group.Segments)
}
if alice.OverlapGroupID != 1 || transcript.Segments[1].OverlapGroupID != 1 {
t.Fatalf("expected final overlap annotation after coalesce: %#v", transcript.Segments)
}
var rpt report.Report
readJSON(t, reportPath, &rpt)
if !hasReportEvent(rpt, "postprocessing", "coalesce", "merged 2 original segment(s) into 1 coalesced segment(s)") {
t.Fatal("expected coalesce report event")
}
}
func TestMergeCoalesceGapOverridePreventsMerge(t *testing.T) {
dir := t.TempDir()
input := writeJSONFile(t, dir, "a.json", `{
"segments": [
{"start": 1, "end": 2, "text": "first"},
{"start": 4, "end": 5, "text": "second"}
]
}`)
output := filepath.Join(dir, "merged.json")
err := executeMerge(
"--input-file", input,
"--output-file", output,
"--coalesce-gap", "1",
)
if err != nil {
t.Fatalf("merge failed: %v", err)
}
var transcript model.FinalTranscript
readJSON(t, output, &transcript)
if len(transcript.Segments) != 2 {
t.Fatalf("segment count = %d, want 2", len(transcript.Segments))
}
if transcript.Segments[0].Text != "first" || transcript.Segments[1].Text != "second" {
t.Fatalf("segments were unexpectedly coalesced: %#v", transcript.Segments)
}
}
func TestSpeakerMatchingUsesFirstMatchingRuleCaseInsensitive(t *testing.T) {
dir := t.TempDir()
input := writeJSONFile(t, dir, "2026-04-19-Adam_Rakestraw.json", `{