narratio/internal/config/load.go

package config

import (
	"fmt"
	"io"
	"os"
	"path/filepath"

	"gopkg.in/yaml.v3"
)

// LoadPipeline loads pipeline configuration from a YAML file with strict field checking.
func LoadPipeline(path string) (*PipelineConfig, error) {
	var cfg PipelineConfig
	if err := decodeStrictYAML("pipeline", path, &cfg); err != nil {
		return nil, fmt.Errorf("load pipeline config: %w", err)
	}
	applyPipelineDefaults(&cfg)
	return &cfg, nil
}

// LoadSession loads session configuration from a YAML file with strict field checking.
func LoadSession(path string) (*SessionConfig, error) {
	var cfg SessionConfig
	if err := decodeStrictYAML("session", path, &cfg); err != nil {
		return nil, fmt.Errorf("load session config: %w", err)
	}
	return &cfg, nil
}

// Load loads and resolves combined pipeline and session configuration.
func Load(pipelinePath, sessionPath string) (*Config, error) {
	pipelineCfg, err := LoadPipeline(pipelinePath)
	if err != nil {
		return nil, err
	}

	sessionCfg, err := LoadSession(sessionPath)
	if err != nil {
		return nil, err
	}

	return &Config{
		Pipeline:     pipelineCfg,
		Session:      sessionCfg,
		PipelinePath: pipelinePath,
		SessionPath:  sessionPath,
	}, nil
}

func decodeStrictYAML(kind, path string, out any) error {
	f, err := os.Open(path)
	if err != nil {
		return fmt.Errorf("%s file %q: open: %w", kind, path, err)
	}
	defer f.Close()

	dec := yaml.NewDecoder(f)
	dec.KnownFields(true)
	if err := dec.Decode(out); err != nil {
		return fmt.Errorf("%s file %q: strict decode failed: %w", kind, path, err)
	}

	var extra any
	if err := dec.Decode(&extra); err != nil && err != io.EOF {
		return fmt.Errorf("%s file %q: trailing content decode failed: %w", kind, path, err)
	}

	return nil
}

func shortName(path, fallback string) string {
	base := filepath.Base(path)
	if base == "." || base == string(filepath.Separator) {
		return fallback
	}
	return base
}

func applyPipelineDefaults(cfg *PipelineConfig) {
	if cfg == nil {
		return
	}
	applyStorageDefaults(&cfg.Storage)
	applySpoolDefaults(&cfg.Spool)
	applyArchiveDefaults(&cfg.Archive)
	applyWhisperXDefaults(&cfg.WhisperX)
	applySeriatimDefaults(&cfg.Seriatim)
	applyAuditaDefaults(&cfg.Audita)
	if cfg.Normalize == nil {
		cfg.Normalize = &NormalizeConfig{}
	}
	applyNormalizeDefaults(cfg.Normalize)
	applyTrimDefaults(cfg.Trim)
	applyScriptoriumDefaults(cfg.Scriptorium)
}

func applyStorageDefaults(cfg *StorageConfig) {
	if cfg == nil {
		return
	}
	if cfg.S3 == nil {
		cfg.S3 = &StorageS3Config{}
	}
	if cfg.S3.RootPrefix == "" {
		cfg.S3.RootPrefix = "dnd"
	}
}

func applySpoolDefaults(cfg *SpoolConfig) {
	if cfg == nil {
		return
	}
	if cfg.Root == "" {
		cfg.Root = "/var/spool/narratio"
	}
}

func applyArchiveDefaults(cfg **ArchiveConfig) {
	if cfg == nil {
		return
	}
	if *cfg == nil {
		*cfg = &ArchiveConfig{}
	}

	if (*cfg).Enabled == nil {
		(*cfg).Enabled = boolPtr(true)
	}
	if (*cfg).UploadRun == nil {
		(*cfg).UploadRun = boolPtr(true)
	}
	if len((*cfg).PromoteArtifacts) == 0 {
		(*cfg).PromoteArtifacts = []ArchivePromotionRule{
			{From: "transcripts/trimmed.json", To: "transcripts/trimmed.json", Required: boolPtr(true)},
			{From: "artifacts/session_recap.md", To: "artifacts/session_recap.md", Required: boolPtr(true)},
		}
	}
	for i := range (*cfg).PromoteArtifacts {
		if (*cfg).PromoteArtifacts[i].Required == nil {
			(*cfg).PromoteArtifacts[i].Required = boolPtr(true)
		}
	}
}

func applyWhisperXDefaults(cfg *WhisperXConfig) {
	if cfg == nil {
		return
	}
	if cfg.Language == "" {
		cfg.Language = "en"
	}
	if cfg.Timeout == "" {
		cfg.Timeout = "30m"
	}
	if cfg.RetryDelay == "" {
		cfg.RetryDelay = "2s"
	}
	if cfg.Concurrency == nil {
		cfg.Concurrency = intPtr(2)
	}
	if cfg.Retries == nil {
		cfg.Retries = intPtr(3)
	}
}

func intPtr(v int) *int {
	p := v
	return &p
}

func applySeriatimDefaults(cfg *SeriatimConfig) {
	if cfg == nil {
		return
	}
	if cfg.Timeout == "" {
		cfg.Timeout = "10m"
	}
	if cfg.OutputSchema == "" {
		cfg.OutputSchema = "seriatim-intermediate"
	}
	if cfg.CoalesceGap == nil {
		cfg.CoalesceGap = float64Ptr(3.0)
	}
	if cfg.Report == nil {
		cfg.Report = boolPtr(true)
	}
}

func applyAuditaDefaults(cfg *AuditaConfig) {
	if cfg == nil {
		return
	}
	if cfg.Timeout == "" {
		cfg.Timeout = "3h"
	}
	if cfg.BaseURL == "" {
		cfg.BaseURL = "https://openrouter.ai/api/v1"
	}
	if cfg.Model == "" {
		cfg.Model = "openrouter/google/gemma-4-31b-it"
	}
	if cfg.Report == nil {
		cfg.Report = boolPtr(true)
	}
}

func applyScriptoriumDefaults(cfg *ScriptoriumConfig) {
	if cfg == nil {
		return
	}
	if cfg.Timeout == "" {
		cfg.Timeout = "10m"
	}
}

func applyTrimDefaults(cfg *TrimConfig) {
	if cfg == nil {
		return
	}
	if cfg.Bounds.Timeout == "" {
		cfg.Bounds.Timeout = "10m"
	}
	if cfg.Seriatim.Report == nil {
		cfg.Seriatim.Report = boolPtr(false)
	}
}

func applyNormalizeDefaults(cfg *NormalizeConfig) {
	if cfg == nil {
		return
	}
	if cfg.OutputSchema == "" {
		cfg.OutputSchema = defaultNormalizeOutputSchema
	}
	if cfg.OutputPath == "" && !cfg.outputPathWasSet() {
		cfg.OutputPath = defaultNormalizeOutputPath
	}
	if cfg.Report == nil {
		cfg.Report = boolPtr(true)
	}
}

func float64Ptr(v float64) *float64 {
	p := v
	return &p
}

func boolPtr(v bool) *bool {
	p := v
	return &p
}