Checkpoint speaker transcriptions incrementally

internal/app/runner_test.go

@@ -1082,7 +1082,7 @@ func TestExecuteStagesReportsSkippedStage(t *testing.T) {
 		t.Fatalf("Save manifest error = %v", err)
 	}
 
-	summary, err := executeStages(context.Background(), cfg, []stage.Stage{BuildFullPlan()[1]}, RunOptions{})
+	summary, err := executeStages(context.Background(), cfg, []stage.Stage{resultStage{name: "transcribe"}}, RunOptions{})
 	if err != nil {
 		t.Fatalf("executeStages() error = %v", err)
 	}
@@ -1302,7 +1302,7 @@ func TestExecuteStagesSkippedStagePreservesExistingOutputsProvenance(t *testing.
 		t.Fatalf("Save manifest error = %v", err)
 	}
 
-	summary, err := executeStages(context.Background(), cfg, []stage.Stage{BuildFullPlan()[1]}, RunOptions{})
+	summary, err := executeStages(context.Background(), cfg, []stage.Stage{resultStage{name: "transcribe"}}, RunOptions{})
 	if err != nil {
 		t.Fatalf("executeStages() error = %v", err)
 	}

internal/app/runner_test_helpers_test.go

@@ -2,10 +2,17 @@ package app
 
 import (
 	"context"
+	"crypto/sha256"
+	"encoding/hex"
+	"os"
+	"path/filepath"
+	"strings"
 	"testing"
 
+	"gitea.maximumdirect.net/eric/narratio/internal/artifacts"
 	"gitea.maximumdirect.net/eric/narratio/internal/config"
 	"gitea.maximumdirect.net/eric/narratio/internal/manifest"
+	"gitea.maximumdirect.net/eric/narratio/internal/pathsafe"
 	"gitea.maximumdirect.net/eric/narratio/internal/stage"
 )
 
@@ -38,9 +45,89 @@ func seedCurrentSemanticEvidence(t *testing.T, cfg *config.Config, m *manifest.M
 			t.Fatalf("fingerprint stage %q: %v", name, err)
 		}
 		record.SemanticConfig = &fingerprint
+		if name == "transcribe" {
+			seedCurrentTranscribeEvidence(t, cfg, m, fingerprint)
+		}
 	}
 }
 
+func seedCurrentTranscribeEvidence(
+	t *testing.T,
+	cfg *config.Config,
+	m *manifest.Manifest,
+	semantic manifest.SemanticConfigFingerprint,
+) {
+	t.Helper()
+	store := artifacts.NewLocalStore(cfg.Pipeline.Workspace.Root)
+	paths := store.SessionPathsFor(cfg.Session.Campaign, m.SessionID)
+	if strings.TrimSpace(m.RunID) == "" {
+		m.RunID = "20260503T010203Z-a1b2c3d4"
+	}
+	records := make(map[string]manifest.TranscribeSpeakerRecord)
+	for _, input := range m.Inputs {
+		if input.Kind != "audio" {
+			continue
+		}
+		audioPath := artifacts.ResolveSessionLocalPathForRead(paths, input.Path)
+		audioBytes, err := os.ReadFile(audioPath)
+		if err != nil {
+			t.Fatalf("read transcribe audio fixture %q: %v", audioPath, err)
+		}
+		audioRelative, err := pathsafe.SlashRelativeFromRoot(paths.Root, audioPath)
+		if err != nil {
+			t.Fatalf("relativize transcribe audio fixture %q: %v", audioPath, err)
+		}
+		speaker := strings.TrimSuffix(filepath.Base(audioPath), filepath.Ext(audioPath))
+		outputRelative := filepath.ToSlash(filepath.Join("transcripts", "raw", speaker+".json"))
+		outputPath, err := pathsafe.JoinSlashRelativeUnderRoot(paths.Root, outputRelative)
+		if err != nil {
+			t.Fatalf("resolve transcribe output fixture: %v", err)
+		}
+		if _, err := os.Stat(outputPath); os.IsNotExist(err) {
+			mustWriteTestFile(t, outputPath, `{"segments":[]}`)
+		} else if err != nil {
+			t.Fatalf("inspect transcribe output fixture %q: %v", outputPath, err)
+		}
+		outputBytes, err := os.ReadFile(outputPath)
+		if err != nil {
+			t.Fatalf("read transcribe output fixture %q: %v", outputPath, err)
+		}
+		audioDigest := sha256.Sum256(audioBytes)
+		outputDigest := sha256.Sum256(outputBytes)
+		fingerprint, err := stage.FingerprintSemanticConfig(
+			manifest.TranscribeFingerprintContractVersion,
+			struct {
+				SpeakerID             string `json:"speaker_id"`
+				AudioPath             string `json:"audio_path"`
+				AudioSHA256           string `json:"audio_sha256"`
+				AudioSize             int64  `json:"audio_size"`
+				SemanticConfigVersion int    `json:"semantic_config_version"`
+				SemanticConfigDigest  string `json:"semantic_config_digest"`
+			}{
+				SpeakerID: speaker, AudioPath: audioRelative,
+				AudioSHA256: hex.EncodeToString(audioDigest[:]), AudioSize: int64(len(audioBytes)),
+				SemanticConfigVersion: semantic.Version, SemanticConfigDigest: semantic.Digest,
+			},
+		)
+		if err != nil {
+			t.Fatalf("fingerprint transcribe speaker fixture %q: %v", speaker, err)
+		}
+		output := manifest.ArtifactRecord{
+			Kind: "transcript_raw", LocalPath: outputRelative,
+			ProducerRunID: m.RunID, Checksum: hex.EncodeToString(outputDigest[:]),
+		}
+		records[speaker] = manifest.TranscribeSpeakerRecord{
+			FingerprintVersion: manifest.TranscribeFingerprintContractVersion,
+			Fingerprint:        fingerprint.Digest, AudioPath: audioRelative,
+			AudioSHA256: hex.EncodeToString(audioDigest[:]), AudioSize: int64(len(audioBytes)),
+			Output: output, OutputSize: int64(len(outputBytes)), ProducerRunID: m.RunID,
+		}
+	}
+	record := m.Stages["transcribe"]
+	record.TranscribeStateVersion = manifest.TranscribeStateContractVersion
+	record.TranscribeSpeakers = records
+}
+
 func loadConfigForSemanticEvidence(t *testing.T, pipelinePath, campaignPath, sessionPath string) *config.Config {
 	t.Helper()
 	cfg, err := config.Load(pipelinePath, campaignPath, sessionPath)