Add campaign configuration support

This commit is contained in:
2026-05-20 20:41:28 -05:00
parent dffb432537
commit b29d8eeb50
34 changed files with 865 additions and 182 deletions

View File

@@ -30,8 +30,10 @@ func TestStagesReturnExpectedMetadata(t *testing.T) {
store := artifacts.NewLocalStore(root)
cfgDir := t.TempDir()
sessionPath := filepath.Join(cfgDir, "session.yml")
campaignPath := filepath.Join(cfgDir, "campaign.yml")
pipelinePath := filepath.Join(cfgDir, "pipeline.yml")
writeStageTestFile(t, sessionPath, "session_id: 2026-05-03\n")
writeStageTestFile(t, campaignPath, "campaign: sample-campaign\ninputs:\n speakers_file: ./speakers.yml\n autocorrect_file: ./autocorrect.yml\n glossary_file: ./glossary.yml\n")
writeStageTestFile(t, pipelinePath, "workspace:\n root: "+root+"\n")
writeStageTestFile(t, filepath.Join(cfgDir, "speakers.yml"), "alice: alice.flac\n")
writeStageTestFile(t, filepath.Join(cfgDir, "autocorrect.yml"), "[]\n")
@@ -48,7 +50,9 @@ func TestStagesReturnExpectedMetadata(t *testing.T) {
env := &Env{
Config: &config.Config{
SessionPath: sessionPath,
CampaignPath: campaignPath,
PipelinePath: pipelinePath,
Campaign: &config.CampaignConfig{Campaign: "sample-campaign"},
Pipeline: &config.PipelineConfig{
Workspace: config.WorkspaceConfig{Root: root},
Storage: config.StorageConfig{
@@ -62,14 +66,28 @@ func TestStagesReturnExpectedMetadata(t *testing.T) {
UploadRun: boolPtr(true),
},
},
StableInputs: config.ResolvedStableInputs{
SpeakersFile: config.ResolvedInputFile{
Path: "./speakers.yml",
ConfigPath: campaignPath,
Source: "campaign_config",
},
AutocorrectFile: config.ResolvedInputFile{
Path: "./autocorrect.yml",
ConfigPath: campaignPath,
Source: "campaign_config",
},
GlossaryFile: config.ResolvedInputFile{
Path: "./glossary.yml",
ConfigPath: campaignPath,
Source: "campaign_config",
},
},
Session: &config.SessionConfig{
SessionID: "2026-05-03",
Campaign: "sample-campaign",
Inputs: config.SessionInputsConfig{
AudioDir: "./audio",
SpeakersFile: "./speakers.yml",
AutocorrectFile: "./autocorrect.yml",
GlossaryFile: "./glossary.yml",
AudioDir: "./audio",
},
},
},

View File

@@ -25,6 +25,7 @@ func (prepareStage) Name() string { return "prepare" }
func (prepareStage) Declares() IODecl {
return IODecl{
Inputs: []artifacts.Ref{
{Kind: "config", Category: "inputs", RelativePath: "campaign.yml"},
{Kind: "config", Category: "inputs", RelativePath: "session.yml"},
{Kind: "config", Category: "inputs", RelativePath: "pipeline.resolved.yml"},
{Kind: "config", Category: "inputs", RelativePath: "speakers.yml"},
@@ -59,21 +60,30 @@ func (prepareStage) Run(ctx context.Context, env *Env, m *manifest.Manifest) (*S
return nil, fmt.Errorf("prepare: ensure workdir layout: %w", err)
}
campaignSrc := env.Config.CampaignPath
if err := requireFile(campaignSrc, "campaign.yml"); err != nil {
return nil, fmt.Errorf("prepare: %w", err)
}
sessionSrc := env.Config.SessionPath
if err := requireFile(sessionSrc, "session.yml"); err != nil {
return nil, fmt.Errorf("prepare: %w", err)
}
sessionDir := filepath.Dir(sessionSrc)
speakersSrc, err := resolvePath(sessionDir, env.Config.Session.Inputs.SpeakersFile)
speakersInput := stableInputSource(env.Config.StableInputs.SpeakersFile, env.Config.Session.Inputs.SpeakersFile, sessionSrc)
autocorrectInput := stableInputSource(env.Config.StableInputs.AutocorrectFile, env.Config.Session.Inputs.AutocorrectFile, sessionSrc)
glossaryInput := stableInputSource(env.Config.StableInputs.GlossaryFile, env.Config.Session.Inputs.GlossaryFile, sessionSrc)
speakersSrc, err := resolveConfigRelativePath(speakersInput)
if err != nil {
return nil, fmt.Errorf("prepare: speakers path: %w", err)
}
autocorrectSrc, err := resolvePath(sessionDir, env.Config.Session.Inputs.AutocorrectFile)
autocorrectSrc, err := resolveConfigRelativePath(autocorrectInput)
if err != nil {
return nil, fmt.Errorf("prepare: autocorrect path: %w", err)
}
glossarySrc, err := resolvePath(sessionDir, env.Config.Session.Inputs.GlossaryFile)
glossarySrc, err := resolveConfigRelativePath(glossaryInput)
if err != nil {
return nil, fmt.Errorf("prepare: glossary path: %w", err)
}
@@ -96,17 +106,27 @@ func (prepareStage) Run(ctx context.Context, env *Env, m *manifest.Manifest) (*S
return nil, fmt.Errorf("prepare: resolve audio inputs: %w", err)
}
inputs := make([]manifest.InputRecord, 0, 5+len(resolvedLocalAudio))
inputs := make([]manifest.InputRecord, 0, 6+len(resolvedLocalAudio))
registerInput := func(kind, path, checksum string) {
inputs = append(inputs, manifest.InputRecord{Kind: kind, Path: path, Checksum: checksum})
}
registerConfigInput := func(kind, path, checksum, source string) {
inputs = append(inputs, manifest.InputRecord{Kind: kind, Path: path, Checksum: checksum, Source: source})
}
campaignDst := filepath.Join(paths.InputsDir, "campaign.yml")
campaignChecksum, err := copyFileIfChanged(env.ArtifactStore, campaignSrc, campaignDst)
if err != nil {
return nil, fmt.Errorf("prepare: materialize campaign.yml: %w", err)
}
registerConfigInput("campaign_config", campaignDst, campaignChecksum, "campaign_config")
sessionDst := filepath.Join(paths.InputsDir, "session.yml")
sessionChecksum, err := copyFileIfChanged(env.ArtifactStore, sessionSrc, sessionDst)
if err != nil {
return nil, fmt.Errorf("prepare: materialize session.yml: %w", err)
}
registerInput("session_config", sessionDst, sessionChecksum)
registerConfigInput("session_config", sessionDst, sessionChecksum, "session_config")
pipelineResolvedBytes, err := renderResolvedPipeline(env.Config.Pipeline)
if err != nil {
@@ -120,19 +140,20 @@ func (prepareStage) Run(ctx context.Context, env *Env, m *manifest.Manifest) (*S
registerInput("pipeline_resolved", pipelineDst, pipelineChecksum)
for _, cfgFile := range []struct {
kind string
src string
dst string
kind string
src string
dst string
source string
}{
{kind: "speakers", src: speakersSrc, dst: filepath.Join(paths.InputsDir, "speakers.yml")},
{kind: "autocorrect", src: autocorrectSrc, dst: filepath.Join(paths.InputsDir, "autocorrect.yml")},
{kind: "glossary", src: glossarySrc, dst: filepath.Join(paths.InputsDir, "glossary.yml")},
{kind: "speakers", src: speakersSrc, dst: filepath.Join(paths.InputsDir, "speakers.yml"), source: speakersInput.Source},
{kind: "autocorrect", src: autocorrectSrc, dst: filepath.Join(paths.InputsDir, "autocorrect.yml"), source: autocorrectInput.Source},
{kind: "glossary", src: glossarySrc, dst: filepath.Join(paths.InputsDir, "glossary.yml"), source: glossaryInput.Source},
} {
checksum, err := copyFileIfChanged(env.ArtifactStore, cfgFile.src, cfgFile.dst)
if err != nil {
return nil, fmt.Errorf("prepare: materialize %s: %w", cfgFile.kind, err)
}
registerInput(cfgFile.kind, cfgFile.dst, checksum)
registerConfigInput(cfgFile.kind, cfgFile.dst, checksum, cfgFile.source)
}
if useS3Audio {
@@ -195,6 +216,31 @@ func renderResolvedPipeline(cfg *config.PipelineConfig) ([]byte, error) {
return yaml.Marshal(cfg)
}
func stableInputSource(resolved config.ResolvedInputFile, fallbackPath, fallbackConfigPath string) config.ResolvedInputFile {
if strings.TrimSpace(resolved.Path) != "" || strings.TrimSpace(resolved.ConfigPath) != "" || strings.TrimSpace(resolved.Source) != "" {
if strings.TrimSpace(resolved.ConfigPath) == "" {
resolved.ConfigPath = fallbackConfigPath
}
if strings.TrimSpace(resolved.Source) == "" {
resolved.Source = "session_config"
}
return resolved
}
return config.ResolvedInputFile{
Path: fallbackPath,
ConfigPath: fallbackConfigPath,
Source: "session_config",
}
}
func resolveConfigRelativePath(input config.ResolvedInputFile) (string, error) {
basePath := strings.TrimSpace(input.ConfigPath)
if basePath == "" {
return "", fmt.Errorf("source config path is required")
}
return resolvePath(filepath.Dir(basePath), input.Path)
}
func resolveAudioInputs(sessionDir string, inputs config.SessionInputsConfig) ([]string, bool, error) {
hasLocal := strings.TrimSpace(inputs.AudioDir) != "" || len(inputs.AudioFiles) > 0
if inputs.AudioS3 != nil {

View File

@@ -35,6 +35,7 @@ func TestPrepareStageExplicitAudioFiles(t *testing.T) {
paths := sessionPathsForEnv(env, m.SessionID)
for _, p := range []string{
filepath.Join(paths.InputsDir, "campaign.yml"),
filepath.Join(paths.InputsDir, "session.yml"),
filepath.Join(paths.InputsDir, "pipeline.resolved.yml"),
filepath.Join(paths.InputsDir, "speakers.yml"),
@@ -48,8 +49,8 @@ func TestPrepareStageExplicitAudioFiles(t *testing.T) {
}
}
if len(m.Inputs) != 7 {
t.Fatalf("manifest inputs len = %d, want 7", len(m.Inputs))
if len(m.Inputs) != 8 {
t.Fatalf("manifest inputs len = %d, want 8", len(m.Inputs))
}
for _, in := range m.Inputs {
if in.Checksum == "" {
@@ -473,8 +474,15 @@ func setupPrepareEnv(t *testing.T) (*Env, *manifest.Manifest) {
sessionPath := filepath.Join(cfgDir, "session.yml")
pipelinePath := filepath.Join(cfgDir, "pipeline.yml")
campaignPath := filepath.Join(cfgDir, "campaign.yml")
writeFile(t, pipelinePath, "workspace:\n root: "+workspace+"\n")
writeFile(t, campaignPath, `campaign: sample-campaign
inputs:
speakers_file: ./speakers.yml
autocorrect_file: ./autocorrect.yml
glossary_file: ./glossary.yml
`)
writeFile(t, sessionPath, "session_id: 2026-05-03\n")
writeFile(t, filepath.Join(cfgDir, "speakers.yml"), "alice: alice.flac\n")
writeFile(t, filepath.Join(cfgDir, "autocorrect.yml"), "[]\n")
@@ -482,16 +490,32 @@ func setupPrepareEnv(t *testing.T) (*Env, *manifest.Manifest) {
cfg := &config.Config{
Pipeline: &config.PipelineConfig{Workspace: config.WorkspaceConfig{Root: workspace}},
Campaign: &config.CampaignConfig{Campaign: "sample-campaign"},
SessionPath: sessionPath,
CampaignPath: campaignPath,
PipelinePath: pipelinePath,
Session: &config.SessionConfig{
SessionID: "2026-05-03",
Campaign: "sample-campaign",
Inputs: config.SessionInputsConfig{
AudioDir: "./audio",
SpeakersFile: "./speakers.yml",
AutocorrectFile: "./autocorrect.yml",
GlossaryFile: "./glossary.yml",
AudioDir: "./audio",
},
},
StableInputs: config.ResolvedStableInputs{
SpeakersFile: config.ResolvedInputFile{
Path: "./speakers.yml",
ConfigPath: campaignPath,
Source: "campaign_config",
},
AutocorrectFile: config.ResolvedInputFile{
Path: "./autocorrect.yml",
ConfigPath: campaignPath,
Source: "campaign_config",
},
GlossaryFile: config.ResolvedInputFile{
Path: "./glossary.yml",
ConfigPath: campaignPath,
Source: "campaign_config",
},
},
}

View File

@@ -233,8 +233,10 @@ func setupTranscribeEnv(t *testing.T, audioFiles []string) (*Env, *manifest.Mani
sessionPath := filepath.Join(cfgDir, "session.yml")
pipelinePath := filepath.Join(cfgDir, "pipeline.yml")
campaignPath := filepath.Join(cfgDir, "campaign.yml")
writeFile(t, sessionPath, "session_id: 2026-05-03\ncampaign: sample-campaign\n")
writeFile(t, pipelinePath, "workspace:\n root: "+workspace+"\n")
writeFile(t, campaignPath, "campaign: sample-campaign\ninputs:\n speakers_file: ./speakers.yml\n autocorrect_file: ./autocorrect.yml\n glossary_file: ./glossary.yml\n")
writeFile(t, filepath.Join(cfgDir, "speakers.yml"), "alice: alice.flac\n")
writeFile(t, filepath.Join(cfgDir, "autocorrect.yml"), "[]\n")
writeFile(t, filepath.Join(cfgDir, "glossary.yml"), "[]\n")
@@ -243,7 +245,9 @@ func setupTranscribeEnv(t *testing.T, audioFiles []string) (*Env, *manifest.Mani
concurrency := 2
cfg := &config.Config{
PipelinePath: pipelinePath,
CampaignPath: campaignPath,
SessionPath: sessionPath,
Campaign: &config.CampaignConfig{Campaign: "sample-campaign"},
Pipeline: &config.PipelineConfig{
Workspace: config.WorkspaceConfig{Root: workspace},
WhisperX: config.WhisperXConfig{
@@ -265,6 +269,11 @@ func setupTranscribeEnv(t *testing.T, audioFiles []string) (*Env, *manifest.Mani
GlossaryFile: "./glossary.yml",
},
},
StableInputs: config.ResolvedStableInputs{
SpeakersFile: config.ResolvedInputFile{Path: "./speakers.yml", ConfigPath: campaignPath, Source: "campaign_config"},
AutocorrectFile: config.ResolvedInputFile{Path: "./autocorrect.yml", ConfigPath: campaignPath, Source: "campaign_config"},
GlossaryFile: config.ResolvedInputFile{Path: "./glossary.yml", ConfigPath: campaignPath, Source: "campaign_config"},
},
}
env := &Env{