sow-tools/internal/topdata/legacy_dataset_migrate.go

package topdata

import (
	"encoding/json"
	"io/fs"
	"os"
	"path/filepath"
	"sort"
	"strings"
)

type legacyDatasetTransform func(relativePath string, obj map[string]any)

func importLegacyDatasetMirror(referenceBuilderDir, dataDir, datasetName, outputName string, transform legacyDatasetTransform) (int, error) {
	legacyDir := filepath.Join(referenceBuilderDir, "data", datasetName)
	if _, err := os.Stat(legacyDir); err != nil {
		if os.IsNotExist(err) {
			return 0, nil
		}
		return 0, err
	}

	targetDir := filepath.Join(dataDir, datasetName)
	baseObj, err := loadJSONObject(filepath.Join(legacyDir, "base.json"))
	if err != nil {
		return 0, err
	}
	if outputName != "" {
		baseObj["output"] = outputName
	}
	if transform != nil {
		transform("base.json", baseObj)
	}

	lockObj, err := loadJSONObject(filepath.Join(legacyDir, "lock.json"))
	if err != nil {
		return 0, err
	}
	if transform != nil {
		transform("lock.json", lockObj)
	}

	moduleRelPaths, err := collectJSONRelativePaths(filepath.Join(legacyDir, "modules"))
	if err != nil {
		return 0, err
	}

	moduleObjs := make(map[string]map[string]any, len(moduleRelPaths))
	for _, relPath := range moduleRelPaths {
		obj, err := loadJSONObject(filepath.Join(legacyDir, "modules", relPath))
		if err != nil {
			return 0, err
		}
		normalizedRel := filepath.ToSlash(relPath)
		if transform != nil {
			transform(filepath.ToSlash(filepath.Join("modules", normalizedRel)), obj)
		}
		moduleObjs[normalizedRel] = obj
	}

	if err := os.MkdirAll(targetDir, 0o755); err != nil {
		return 0, err
	}

	updated := 0
	for _, write := range []struct {
		path string
		obj  map[string]any
	}{
		{path: filepath.Join(targetDir, "base.json"), obj: baseObj},
		{path: filepath.Join(targetDir, "lock.json"), obj: lockObj},
	} {
		changed, err := writeJSONObjectIfChanged(write.path, write.obj)
		if err != nil {
			return 0, err
		}
		if changed {
			updated++
		}
	}

	targetModulesDir := filepath.Join(targetDir, "modules")
	if len(moduleObjs) > 0 {
		if err := os.MkdirAll(targetModulesDir, 0o755); err != nil {
			return 0, err
		}
	}
	for relPath, obj := range moduleObjs {
		targetPath := filepath.Join(targetModulesDir, filepath.FromSlash(relPath))
		if err := os.MkdirAll(filepath.Dir(targetPath), 0o755); err != nil {
			return 0, err
		}
		changed, err := writeJSONObjectIfChanged(targetPath, obj)
		if err != nil {
			return 0, err
		}
		if changed {
			updated++
		}
	}

	staleModules, err := collectJSONRelativePaths(targetModulesDir)
	if err != nil {
		return 0, err
	}
	for _, relPath := range staleModules {
		normalizedRel := filepath.ToSlash(relPath)
		if _, ok := moduleObjs[normalizedRel]; ok {
			continue
		}
		if err := os.Remove(filepath.Join(targetModulesDir, filepath.FromSlash(normalizedRel))); err != nil && !os.IsNotExist(err) {
			return 0, err
		}
		updated++
	}

	return updated, nil
}

func collectJSONRelativePaths(root string) ([]string, error) {
	if _, err := os.Stat(root); err != nil {
		if os.IsNotExist(err) {
			return nil, nil
		}
		return nil, err
	}

	var relPaths []string
	err := filepath.WalkDir(root, func(path string, d fs.DirEntry, err error) error {
		if err != nil {
			return err
		}
		if d.IsDir() {
			return nil
		}
		if !strings.EqualFold(filepath.Ext(d.Name()), ".json") {
			return nil
		}
		relPath, err := filepath.Rel(root, path)
		if err != nil {
			return err
		}
		relPaths = append(relPaths, filepath.ToSlash(relPath))
		return nil
	})
	if err != nil {
		return nil, err
	}
	sort.Strings(relPaths)
	return relPaths, nil
}

func writeJSONObjectIfChanged(path string, obj map[string]any) (bool, error) {
	raw, err := json.MarshalIndent(obj, "", "  ")
	if err != nil {
		return false, err
	}
	raw = append(raw, '\n')

	current, err := os.ReadFile(path)
	if err == nil && string(current) == string(raw) {
		return false, nil
	}
	if err != nil && !os.IsNotExist(err) {
		return false, err
	}
	if err := os.WriteFile(path, raw, 0o644); err != nil {
		return false, err
	}
	return true, nil
}

func fileExists(path string) bool {
	_, err := os.Stat(path)
	return err == nil
}