summercms/phrasebook/loader.go

package phrasebook

import (
	"fmt"
	"io/fs"
	"path"
	"sort"
	"strings"
	"unicode"

	"github.com/goccy/go-yaml"
)

// Catalog is an immutable-after-load set of namespaced translation entries.
type Catalog struct {
	// locale -> full key (vendor.plugin::group.dot.path) -> entry
	messages map[string]map[string]entry
	// namespace (plugin ID) -> owning plugin ID
	namespaces map[string]string
}

type entry struct {
	text string
}

// NewCatalog returns an empty translation catalog.
func NewCatalog() *Catalog {
	return &Catalog{
		messages:   make(map[string]map[string]entry),
		namespaces: make(map[string]string),
	}
}

// Load walks fsys for lang/<locale>/<group>.yaml files, flattens nested YAML
// maps into vendor.plugin::group.dot.path keys, and merges them into c.
func (c *Catalog) Load(pluginID string, fsys fs.FS) error {
	if c == nil {
		return fmt.Errorf("phrasebook: catalog is nil")
	}
	if pluginID == "" {
		return fmt.Errorf("phrasebook: plugin id is empty")
	}
	if fsys == nil {
		return nil
	}
	if owner, ok := c.namespaces[pluginID]; ok {
		return catalogErrorf(pluginID, "", "", "duplicate namespace owner %q", owner)
	}
	if c.messages == nil {
		c.messages = make(map[string]map[string]entry)
	}
	if c.namespaces == nil {
		c.namespaces = make(map[string]string)
	}

	var files []string
	err := fs.WalkDir(fsys, ".", func(p string, d fs.DirEntry, err error) error {
		if err != nil {
			return err
		}
		if d.IsDir() {
			return nil
		}
		files = append(files, path.Clean(p))
		return nil
	})
	if err != nil {
		return catalogErrorf(pluginID, "", "", "walk lang fs: %s", err.Error())
	}
	sort.Strings(files)

	for _, file := range files {
		locale, group, err := parseLangPath(file)
		if err != nil {
			return catalogErrorf(pluginID, file, "", "%s", err.Error())
		}
		raw, err := fs.ReadFile(fsys, file)
		if err != nil {
			return catalogErrorf(pluginID, file, "", "read: %s", err.Error())
		}
		if err := c.loadFile(pluginID, file, locale, group, raw); err != nil {
			return err
		}
	}
	c.namespaces[pluginID] = pluginID
	return nil
}

func (c *Catalog) loadFile(pluginID, file, locale, group string, raw []byte) error {
	var decoded any
	if err := yaml.Unmarshal(raw, &decoded); err != nil {
		return catalogErrorf(pluginID, file, "", "parse: %s", err.Error())
	}
	if decoded == nil {
		return nil
	}
	root, err := asStringMap(decoded)
	if err != nil {
		return catalogErrorf(pluginID, file, "", "%s", err.Error())
	}
	prefix := pluginID + "::" + group
	return flatten(c, pluginID, file, locale, prefix, root)
}

func flatten(c *Catalog, pluginID, file, locale, prefix string, v any) error {
	switch val := v.(type) {
	case string:
		return c.put(pluginID, file, locale, prefix, val)
	case map[string]any:
		if len(val) == 0 {
			return nil
		}
		keys := make([]string, 0, len(val))
		for k := range val {
			keys = append(keys, k)
		}
		sort.Strings(keys)
		for _, k := range keys {
			next := joinKey(prefix, k)
			if err := flatten(c, pluginID, file, locale, next, val[k]); err != nil {
				return err
			}
		}
		return nil
	case map[any]any:
		converted, err := anyMapToStringMap(val)
		if err != nil {
			return catalogErrorf(pluginID, file, prefix, "%s", err.Error())
		}
		return flatten(c, pluginID, file, locale, prefix, converted)
	case nil:
		return catalogErrorf(pluginID, file, prefix, "non-string leaf (null)")
	default:
		return catalogErrorf(pluginID, file, prefix, "non-string leaf (%T)", v)
	}
}

func (c *Catalog) put(pluginID, file, locale, key, text string) error {
	if key == "" || strings.HasSuffix(key, "::") || strings.HasSuffix(key, ".") {
		return catalogErrorf(pluginID, file, key, "empty key")
	}
	byKey := c.messages[locale]
	if byKey == nil {
		byKey = make(map[string]entry)
		c.messages[locale] = byKey
	}
	if _, exists := byKey[key]; exists {
		return catalogErrorf(pluginID, file, key, "duplicate key")
	}
	byKey[key] = entry{text: text}
	return nil
}

func (c *Catalog) lookup(locale, key string) (entry, bool) {
	if c == nil {
		return entry{}, false
	}
	byKey := c.messages[locale]
	if byKey == nil {
		return entry{}, false
	}
	e, ok := byKey[key]
	return e, ok
}

func parseLangPath(p string) (locale, group string, err error) {
	p = path.Clean("/" + strings.ReplaceAll(p, "\\", "/"))
	p = strings.TrimPrefix(p, "/")
	if p == ".." || strings.HasPrefix(p, "../") || strings.Contains(p, "/../") {
		return "", "", fmt.Errorf("malformed path")
	}
	parts := strings.Split(p, "/")
	if len(parts) != 3 || parts[0] != "lang" {
		return "", "", fmt.Errorf("malformed path")
	}
	locale = parts[1]
	base := parts[2]
	if !strings.HasSuffix(base, ".yaml") {
		return "", "", fmt.Errorf("malformed path")
	}
	group = strings.TrimSuffix(base, ".yaml")
	if !validLocale(locale) || !validGroup(group) {
		return "", "", fmt.Errorf("malformed path")
	}
	return locale, group, nil
}

func validLocale(s string) bool {
	if s == "" || s == "." || s == ".." {
		return false
	}
	for i, r := range s {
		if unicode.IsLetter(r) || unicode.IsDigit(r) {
			continue
		}
		if (r == '-' || r == '_') && i > 0 && i < len(s)-1 {
			continue
		}
		return false
	}
	return true
}

func validGroup(s string) bool {
	if s == "" || s == "." || s == ".." {
		return false
	}
	for _, r := range s {
		if unicode.IsLetter(r) || unicode.IsDigit(r) || r == '_' || r == '-' {
			continue
		}
		return false
	}
	return true
}

func joinKey(prefix, key string) string {
	if prefix == "" {
		return key
	}
	if key == "" {
		return prefix
	}
	return prefix + "." + key
}

func asStringMap(v any) (map[string]any, error) {
	switch val := v.(type) {
	case map[string]any:
		return val, nil
	case map[any]any:
		return anyMapToStringMap(val)
	default:
		return nil, fmt.Errorf("root must be a mapping")
	}
}

func anyMapToStringMap(v map[any]any) (map[string]any, error) {
	out := make(map[string]any, len(v))
	for k, val := range v {
		sk, ok := k.(string)
		if !ok {
			return nil, fmt.Errorf("non-string key (%T)", k)
		}
		out[sk] = val
	}
	return out, nil
}

func catalogErrorf(pluginID, file, key, format string, args ...any) error {
	msg := fmt.Sprintf(format, args...)
	switch {
	case key != "":
		return fmt.Errorf("phrasebook: plugin %s file %s key %s: %s", pluginID, file, key, msg)
	case file != "":
		return fmt.Errorf("phrasebook: plugin %s file %s: %s", pluginID, file, msg)
	default:
		return fmt.Errorf("phrasebook: plugin %s: %s", pluginID, msg)
	}
}