Add output validation against a defined JSON schema

internal/pipeline/runner.go

@@ -4,14 +4,16 @@ import (
 	"context"
 	"fmt"
 
+	"gitea.maximumdirect.net/eric/seriatim/internal/artifact"
 	"gitea.maximumdirect.net/eric/seriatim/internal/config"
 	"gitea.maximumdirect.net/eric/seriatim/internal/model"
 	"gitea.maximumdirect.net/eric/seriatim/internal/report"
+	"gitea.maximumdirect.net/eric/seriatim/schema"
 )
 
 const (
-	applicationName = "seriatim"
-	version         = "dev"
+	applicationName = artifact.ApplicationName
+	version         = artifact.Version
 )
 
 // Run validates module composition, executes the pipeline, and emits outputs.
@@ -139,29 +141,8 @@ func validatePreprocessors(modules []Preprocessor) error {
 	return nil
 }
 
-func finalizeTranscript(cfg config.Config, merged model.MergedTranscript) model.FinalTranscript {
-	segments := make([]model.Segment, len(merged.Segments))
-	copy(segments, merged.Segments)
-	for index := range segments {
-		segments[index].Words = nil
-		segments[index].DerivedFrom = append([]string(nil), segments[index].DerivedFrom...)
-	}
-	overlapGroups := make([]model.OverlapGroup, len(merged.OverlapGroups))
-	copy(overlapGroups, merged.OverlapGroups)
-
-	return model.FinalTranscript{
-		Metadata: model.OutputMetadata{
-			Application:           applicationName,
-			Version:               version,
-			InputReader:           cfg.InputReader,
-			InputFiles:            append([]string(nil), cfg.InputFiles...),
-			PreprocessingModules:  append([]string(nil), cfg.PreprocessingModules...),
-			PostprocessingModules: append([]string(nil), cfg.PostprocessingModules...),
-			OutputModules:         append([]string(nil), cfg.OutputModules...),
-		},
-		Segments:      segments,
-		OverlapGroups: overlapGroups,
-	}
+func finalizeTranscript(cfg config.Config, merged model.MergedTranscript) schema.Transcript {
+	return artifact.FromMerged(cfg, merged)
 }
 
 func finalizeReport(cfg config.Config, events []report.Event) report.Report {