Files
narratio/internal/config/config.go

390 lines
18 KiB
Go

package config
import configkit "gitea.maximumdirect.net/eric/ConfigKit"
// Config is the resolved combined configuration from pipeline.yml,
// campaign.yml, and session.yml.
type Config struct {
Pipeline *PipelineConfig
Campaign *CampaignConfig
Session *SessionConfig
PipelinePath string
CampaignPath string
SessionPath string
StableInputs ResolvedStableInputs
SessionSource SessionSource
Party ResolvedParty
}
// PipelineConfig contains durable pipeline-level settings.
type PipelineConfig struct {
Workspace WorkspaceConfig `yaml:"workspace" json:"workspace"`
Campaigns CampaignsConfig `yaml:"campaigns" json:"campaigns"`
Storage StorageConfig `yaml:"storage" json:"storage"`
Spool SpoolConfig `yaml:"spool" json:"spool"`
Cache CacheConfig `yaml:"cache" json:"cache"`
Publish *PublishConfig `yaml:"publish" json:"publish"`
WhisperX WhisperXConfig `yaml:"whisperx" json:"whisperx"`
Seriatim SeriatimConfig `yaml:"seriatim" json:"seriatim"`
Audita AuditaConfig `yaml:"audita" json:"audita"`
Normalize *NormalizeConfig `yaml:"normalize" json:"normalize"`
Trim *TrimConfig `yaml:"trim" json:"trim"`
Render *RenderConfig `yaml:"render" json:"render"`
Scriptorium *ScriptoriumConfig `yaml:"scriptorium" json:"scriptorium"`
Notarius *NotariusConfig `yaml:"notarius" json:"notarius"`
Notification NotificationConfig `yaml:"notification" json:"notification"`
resolution *pipelineResolutionMetadata `yaml:"-" json:"-"`
familyCatalog ArtifactFamilyCatalog `yaml:"-" json:"-"`
}
// ArtifactFamilies returns a copy of runtime-only expansion provenance.
func ArtifactFamilies(cfg *PipelineConfig) ArtifactFamilyCatalog {
if cfg == nil {
return ArtifactFamilyCatalog{}
}
return cloneArtifactFamilyCatalog(cfg.familyCatalog)
}
// CampaignsConfig configures the local campaign registry.
type CampaignsConfig struct {
Root string `yaml:"root" json:"root"`
DefaultCampaignID string `yaml:"default_campaign_id" json:"default_campaign_id"`
}
// CampaignConfig contains stable campaign-level identity and input defaults.
type CampaignConfig struct {
CampaignID string `yaml:"campaign_id" json:"campaign_id"`
SessionTemplateFile string `yaml:"session_template_file" json:"session_template_file"`
Inputs CampaignInputsConfig `yaml:"inputs" json:"inputs"`
}
// CampaignInputsConfig contains stable campaign-level input file references.
type CampaignInputsConfig struct {
SpeakersFile string `yaml:"speakers_file" json:"speakers_file"`
AutocorrectFile string `yaml:"autocorrect_file" json:"autocorrect_file"`
GlossaryFile string `yaml:"glossary_file" json:"glossary_file"`
PartyFile string `yaml:"party_file" json:"party_file"`
SpellCatalogFile string `yaml:"spell_catalog_file" json:"spell_catalog_file"`
}
// SessionConfig contains per-session inputs and metadata.
type SessionConfig struct {
SessionID string `yaml:"session_id" json:"session_id"`
PreviousSessionID string `yaml:"previous_session_id" json:"previous_session_id"`
Campaign string `yaml:"campaign" json:"campaign"`
Date string `yaml:"date" json:"date"`
Title string `yaml:"title" json:"title"`
Inputs SessionInputsConfig `yaml:"inputs" json:"inputs"`
}
// WorkspaceConfig configures local workspace behavior.
type WorkspaceConfig struct {
Root string `yaml:"root" json:"root"`
CleanupAfterPublish bool `yaml:"cleanup_after_publish" json:"cleanup_after_publish"`
}
// StorageConfig configures storage backends and related parameters.
type StorageConfig struct {
Backend string `yaml:"backend" json:"backend"`
S3 *StorageS3Config `yaml:"s3" json:"s3"`
}
const (
// StorageBackendLocal disables remote object-store operations.
StorageBackendLocal = "local"
// StorageBackendS3 enables the configured S3 object store.
StorageBackendS3 = "s3"
)
// StorageS3Config configures S3 storage coordinates.
type StorageS3Config struct {
Bucket string `yaml:"bucket" json:"bucket"`
RootPrefix string `yaml:"root_prefix" json:"root_prefix"`
Region string `yaml:"region" json:"region"`
Endpoint string `yaml:"endpoint" json:"endpoint"`
ForcePathStyle bool `yaml:"force_path_style" json:"force_path_style"`
AccessKeyID *configkit.Secret `yaml:"access_key_id,omitempty" json:"access_key_id,omitempty"`
SecretAccessKey *configkit.Secret `yaml:"secret_access_key,omitempty" json:"secret_access_key,omitempty"`
}
// SpoolConfig configures local spool storage for staged data.
type SpoolConfig struct {
Root string `yaml:"root" json:"root"`
DeleteAudioAfterPublish bool `yaml:"delete_audio_after_publish" json:"delete_audio_after_publish"`
}
// CacheConfig configures durable local caches for reusable remote inputs.
type CacheConfig struct {
Root string `yaml:"root" json:"root"`
S3Audio *bool `yaml:"s3_audio" json:"s3_audio"`
}
// PublishConfig configures publish behavior and source-based output uploads.
type PublishConfig struct {
Enabled *bool `yaml:"enabled" json:"enabled"`
UploadRun *bool `yaml:"upload_run" json:"upload_run"`
Outputs []PublishOutputRule `yaml:"outputs" json:"outputs"`
}
// PublishOutputRule configures one source-to-destination publish mapping.
type PublishOutputRule struct {
Source string `yaml:"source" json:"source"`
Dest string `yaml:"dest" json:"dest"`
Required *bool `yaml:"required" json:"required"`
}
// WhisperXConfig configures WhisperX adapter settings.
type WhisperXConfig struct {
TranscribeURL string `yaml:"transcribe_url" json:"transcribe_url"`
Language string `yaml:"language" json:"language"`
Timeout string `yaml:"timeout" json:"timeout"`
Retries *int `yaml:"retries" json:"retries"`
RetryDelay string `yaml:"retry_delay" json:"retry_delay"`
Concurrency *int `yaml:"concurrency" json:"concurrency"`
}
// SeriatimConfig configures seriatim adapter settings.
type SeriatimConfig struct {
Binary string `yaml:"binary" json:"binary"`
Timeout string `yaml:"timeout" json:"timeout"`
OutputSchema string `yaml:"output_schema" json:"output_schema"`
CoalesceGap *float64 `yaml:"coalesce_gap" json:"coalesce_gap"`
Report *bool `yaml:"report" json:"report"`
Env SeriatimEnvConfig `yaml:"env" json:"env"`
}
// SeriatimEnvConfig configures optional advanced seriatim environment tuning.
type SeriatimEnvConfig struct {
OverlapWordRunGap *float64 `yaml:"overlap_word_run_gap" json:"overlap_word_run_gap"`
OverlapWordRunReorderWindow *float64 `yaml:"overlap_word_run_reorder_window" json:"overlap_word_run_reorder_window"`
BackchannelMaxDuration *float64 `yaml:"backchannel_max_duration" json:"backchannel_max_duration"`
FillerMaxDuration *float64 `yaml:"filler_max_duration" json:"filler_max_duration"`
}
// AuditaConfig configures audita adapter settings.
type AuditaConfig struct {
Binary string `yaml:"binary" json:"binary"`
Timeout string `yaml:"timeout" json:"timeout"`
LLMAPIKey *configkit.Secret `yaml:"llm_api_key,omitempty" json:"llm_api_key,omitempty"`
Modules []string `yaml:"modules" json:"modules"`
BaseURL string `yaml:"base_url" json:"base_url"`
Model string `yaml:"model" json:"model"`
TotalLLMConcurrency *int `yaml:"total_llm_concurrency" json:"total_llm_concurrency"`
ProposalLLMConcurrency *int `yaml:"proposal_llm_concurrency" json:"proposal_llm_concurrency"`
ValidationModel string `yaml:"validation_model" json:"validation_model"`
ValidationLLMConcurrency *int `yaml:"validation_llm_concurrency" json:"validation_llm_concurrency"`
TranscriptDescription string `yaml:"transcript_description" json:"transcript_description"`
ConfigPath string `yaml:"config_path" json:"config_path"`
OutputSchema string `yaml:"output_schema" json:"output_schema"`
WorkDirRetention string `yaml:"work_dir_retention" json:"work_dir_retention"`
Report *bool `yaml:"report" json:"report"`
}
// NormalizeConfig configures normalize-stage transcript schema/output behavior.
type NormalizeConfig struct {
OutputPath string `yaml:"output_path" json:"output_path"`
OutputSchema string `yaml:"output_schema" json:"output_schema"`
Report *bool `yaml:"report" json:"report"`
}
// TrimConfig configures trim-stage transcript boundary behavior.
type TrimConfig struct {
Enabled *bool `yaml:"enabled" json:"enabled"`
OutputPath string `yaml:"output_path" json:"output_path"`
Bounds TrimBoundsConfig `yaml:"bounds" json:"bounds"`
Seriatim TrimSeriatimConfig `yaml:"seriatim" json:"seriatim"`
}
// TrimBoundsConfig configures Scriptorium bounds prompt execution for trim.
type TrimBoundsConfig struct {
PromptID string `yaml:"prompt_id" json:"prompt_id"`
ProfileID string `yaml:"profile_id" json:"profile_id"`
TranscriptInputName string `yaml:"transcript_input_name" json:"transcript_input_name"`
OutputPath string `yaml:"output_path" json:"output_path"`
Timeout string `yaml:"timeout" json:"timeout"`
RenderDebug bool `yaml:"render_debug" json:"render_debug"`
RenderOutputPath string `yaml:"render_output_path" json:"render_output_path"`
}
// TrimSeriatimConfig configures Seriatim options for trim output construction.
type TrimSeriatimConfig struct {
Report *bool `yaml:"report" json:"report"`
}
// RenderConfig configures render-stage output formatting behavior.
type RenderConfig struct {
Enabled *bool `yaml:"enabled" json:"enabled"`
Format string `yaml:"format" json:"format"`
Title string `yaml:"title" json:"title"`
IncludeTimestamps *bool `yaml:"include_timestamps" json:"include_timestamps"`
IncludeSegmentIDs *bool `yaml:"include_segment_ids" json:"include_segment_ids"`
IncludeMetadata bool `yaml:"include_metadata" json:"include_metadata"`
}
// ScriptoriumConfig configures Scriptorium-backed artifact generation.
type ScriptoriumConfig struct {
Binary string `yaml:"binary" json:"binary"`
ConfigPath string `yaml:"config_path" json:"config_path"`
Timeout string `yaml:"timeout" json:"timeout"`
RenderDebug bool `yaml:"render_debug" json:"render_debug"`
Artifacts map[string]ScriptoriumArtifactConfig `yaml:"artifacts" json:"artifacts"`
ArtifactFamilies map[string]ScriptoriumArtifactFamilyConfig `yaml:"artifact_families" json:"artifact_families"`
}
// ScriptoriumArtifactConfig configures one named output artifact workflow.
type ScriptoriumArtifactConfig struct {
Enabled bool `yaml:"enabled" json:"enabled"`
DependsOn []string `yaml:"depends_on" json:"depends_on"`
RenderDebug *bool `yaml:"render_debug" json:"render_debug"`
PromptID string `yaml:"prompt_id" json:"prompt_id"`
ProfileID string `yaml:"profile_id" json:"profile_id"`
OutputPath string `yaml:"output_path" json:"output_path"`
Timeout string `yaml:"timeout" json:"timeout"`
Inputs map[string]ScriptoriumInputConfig `yaml:"inputs" json:"inputs"`
Vars map[string]any `yaml:"vars" json:"vars"`
}
// ScriptoriumArtifactFamilyConfig declares the shared configuration expanded
// into one ordinary artifact for each canonical party character.
type ScriptoriumArtifactFamilyConfig struct {
Enabled bool `yaml:"enabled" json:"enabled"`
ForEach string `yaml:"for_each" json:"for_each"`
DependsOn []string `yaml:"depends_on" json:"depends_on"`
RenderDebug *bool `yaml:"render_debug" json:"render_debug"`
PromptID string `yaml:"prompt_id" json:"prompt_id"`
ProfileID string `yaml:"profile_id" json:"profile_id"`
OutputPathPattern string `yaml:"output_path_pattern" json:"output_path_pattern"`
Timeout string `yaml:"timeout" json:"timeout"`
Inputs map[string]ScriptoriumInputConfig `yaml:"inputs" json:"inputs"`
Vars map[string]any `yaml:"vars" json:"vars"`
MemberVars map[string]string `yaml:"member_vars" json:"member_vars"`
MemberDependencies []string `yaml:"member_dependencies" json:"member_dependencies"`
Publish *ScriptoriumArtifactFamilyPublishConfig `yaml:"publish" json:"publish"`
}
// ScriptoriumArtifactFamilyPublishConfig retains the typed declaration for
// family publishing. It is resolved by the publish owner.
type ScriptoriumArtifactFamilyPublishConfig struct {
Enabled bool `yaml:"enabled" json:"enabled"`
Required bool `yaml:"required" json:"required"`
DestPattern string `yaml:"dest_pattern" json:"dest_pattern"`
}
// ArtifactFamilyCatalog records the runtime-only origin of expanded artifacts.
// It is intentionally kept outside the Scriptorium adapter configuration.
type ArtifactFamilyCatalog struct {
Families map[string]ArtifactFamilyOrigin
Members map[string]ArtifactFamilyMemberOrigin
}
// ArtifactFamilyOrigin records one declared family and its source ownership.
type ArtifactFamilyOrigin struct {
Members []string
MemberDependencies []string
Publish *ScriptoriumArtifactFamilyPublishConfig
Source string
}
// ArtifactFamilyMemberOrigin identifies the canonical party member that
// produced one ordinary concrete artifact.
type ArtifactFamilyMemberOrigin struct {
Family string
CharacterID string
Source string
Dependencies []string
Inputs map[string]string
}
// ScriptoriumInputConfig configures one named prompt input source.
type ScriptoriumInputConfig struct {
Source string `yaml:"source" json:"source"`
Required bool `yaml:"required" json:"required"`
}
// NotariusConfig configures structured artifact extraction by Notarius.
type NotariusConfig struct {
Enabled bool `yaml:"enabled" json:"enabled"`
Binary string `yaml:"binary" json:"binary"`
ConfigPath string `yaml:"config_path" json:"config_path"`
PipelineID string `yaml:"pipeline_id" json:"pipeline_id"`
Timeout string `yaml:"timeout" json:"timeout"`
WorkingDirectory string `yaml:"working_directory" json:"working_directory"`
References map[string]string `yaml:"references" json:"references"`
Outputs map[string]NotariusOutputConfig `yaml:"outputs" json:"outputs"`
}
// NotariusOutputConfig declares one required extraction lane contract.
type NotariusOutputConfig struct {
LaneID string `yaml:"lane_id" json:"lane_id"`
MediaType string `yaml:"media_type" json:"media_type"`
SchemaID string `yaml:"schema_id" json:"schema_id"`
SchemaVersion string `yaml:"schema_version" json:"schema_version"`
ModuleKey string `yaml:"module_key" json:"module_key"`
}
// NotificationConfig configures the supported notification behavior.
type NotificationConfig struct {
Mode string `yaml:"mode" json:"mode"`
}
// SessionInputsConfig contains per-session input references.
type SessionInputsConfig struct {
AudioDir string `yaml:"audio_dir" json:"audio_dir"`
AudioFiles []string `yaml:"audio_files" json:"audio_files"`
AudioS3 *SessionAudioS3Input `yaml:"audio_s3" json:"audio_s3"`
SpeakersFile string `yaml:"speakers_file" json:"speakers_file"`
AutocorrectFile string `yaml:"autocorrect_file" json:"autocorrect_file"`
GlossaryFile string `yaml:"glossary_file" json:"glossary_file"`
SpellCatalogFile string `yaml:"spell_catalog_file" json:"spell_catalog_file"`
}
// SessionAudioS3Input configures S3 session-audio input discovery.
type SessionAudioS3Input struct {
Prefix string `yaml:"prefix" json:"prefix"`
}
// ResolvedStableInputs records where stable input file paths came from after
// campaign/session merge.
type ResolvedStableInputs struct {
SpeakersFile ResolvedInputFile
AutocorrectFile ResolvedInputFile
GlossaryFile ResolvedInputFile
PartyFile ResolvedInputFile
SpellCatalogFile ResolvedInputFile
}
// ResolvedInputFile records one merged config path and its source config file.
type ResolvedInputFile struct {
Path string
ConfigPath string
Source string
}
// ResolvedParty records the campaign-owned canonical party and its non-secret
// source provenance. It is runtime-only and must never be copied into manifests
// as raw party content.
type ResolvedParty struct {
Source PartySource
Canonical *CanonicalParty
}
// PartySource identifies the selected party input without retaining its raw
// contents. Path is the resolved on-disk source path.
type PartySource struct {
Path string
ConfigPath string
Source string
}
// SessionSource records where session.yml came from before materialization.
type SessionSource struct {
Source string
LocalPath string
S3Bucket string
S3Key string
S3Size int64
S3ETag string
}