sm-user-plugin/classes/handle_generator.go

package classes

import (
	"context"
	"errors"
	"strconv"
	"strings"
	"unicode"

	"git.golem15.com/golem15/sm-user-plugin/models"
	"gorm.io/gorm"
)

const (
	handleMinLen = 2
	handleMaxLen = 255
)

// HandleGenerator ports the PHP ChatChannels HandleGenerator. The emitted
// charset [a-z0-9._-] is a strict subset of MentionParser [A-Za-z0-9._-],
// length 2–255, so every generated handle is matchable as an @-token.
type HandleGenerator struct{}

// Slug transliterates name to ASCII, lowercases, drops every rune outside
// [a-z0-9._-], then trims leading/trailing `._-`. A result shorter than 2
// is not a valid base (Generate falls back to user{id}).
func (HandleGenerator) Slug(name string) string {
	name = strings.TrimSpace(name)
	if name == "" {
		return ""
	}
	var b strings.Builder
	b.Grow(len(name))
	for _, r := range name {
		b.WriteString(foldHandleRune(r))
	}
	return strings.Trim(b.String(), "._-")
}

// Generate returns a unique handle for name. The base is Slug(name); if
// that is shorter than 2 it falls back to "user{fallbackID}" (or "user"
// when no id is known). The smallest free integer suffix is then appended
// until isTaken returns false: first "James" → "james", second → "james1".
func (g HandleGenerator) Generate(name string, isTaken func(string) bool, fallbackID *uint) string {
	base := g.Slug(name)
	if len(base) < handleMinLen {
		if fallbackID != nil {
			base = "user" + strconv.FormatUint(uint64(*fallbackID), 10)
		} else {
			base = "user"
		}
	}
	base = clipHandle(base)
	candidate := base
	i := 0
	for isTaken != nil && isTaken(candidate) {
		i++
		candidate = clipHandle(base + strconv.Itoa(i))
	}
	return candidate
}

// AssignHandle writes a unique handle onto user when it has none, using
// the user's display name and id. It is idempotent when handle is already
// set. The write uses db (the create transaction when there is one).
func AssignHandle(ctx context.Context, db *gorm.DB, user *models.User) error {
	if user == nil {
		return errors.New("golem15.user: a handle needs a user")
	}
	if user.Handle != nil && strings.TrimSpace(*user.Handle) != "" {
		return nil
	}
	if db == nil {
		return errors.New("golem15.user: a handle needs a database")
	}
	name := ""
	if user.Name != nil {
		name = *user.Name
	}
	var fallback *uint
	if user.ID != 0 {
		id := user.ID
		fallback = &id
	}
	gen := HandleGenerator{}
	for attempt := 0; attempt < 8; attempt++ {
		handle := gen.Generate(name, func(candidate string) bool {
			var n int64
			q := db.WithContext(ctx).Model(&models.User{}).Where("handle = ?", candidate)
			if user.ID != 0 {
				q = q.Where("id <> ?", user.ID)
			}
			if err := q.Count(&n).Error; err != nil {
				return true
			}
			return n > 0
		}, fallback)
		user.Handle = &handle
		res := db.WithContext(ctx).Model(user).Where("id = ?", user.ID).Update("handle", handle)
		if res.Error == nil {
			return nil
		}
		if !isUniqueHandleConflict(res.Error) {
			return res.Error
		}
	}
	return errors.New("golem15.user: could not allocate a unique handle")
}

func clipHandle(s string) string {
	if len(s) <= handleMaxLen {
		return s
	}
	return s[:handleMaxLen]
}

func isUniqueHandleConflict(err error) bool {
	if err == nil {
		return false
	}
	if errors.Is(err, gorm.ErrDuplicatedKey) {
		return true
	}
	msg := strings.ToLower(err.Error())
	return strings.Contains(msg, "users_handle_unique") || strings.Contains(msg, "duplicate key")
}

// foldHandleRune maps r into the handle charset [a-z0-9._-] or "".
func foldHandleRune(r rune) string {
	r = unicode.ToLower(r)
	switch {
	case r >= 'a' && r <= 'z', r >= '0' && r <= '9', r == '.', r == '_', r == '-':
		return string(r)
	}
	if mapped, ok := latinHandleFold[r]; ok {
		return mapped
	}
	return ""
}

// latinHandleFold covers Latin-1 / Latin Extended-A letters PHP Str::ascii
// would emit for the names we generate handles from (Polish, French, German).
var latinHandleFold = map[rune]string{
	'à': "a", 'á': "a", 'â': "a", 'ã': "a", 'ä': "a", 'å': "a", 'ą': "a", 'ā': "a",
	'æ': "ae",
	'ç': "c", 'ć': "c", 'č': "c",
	'ď': "d", 'đ': "d", 'ð': "d",
	'è': "e", 'é': "e", 'ê': "e", 'ë': "e", 'ę': "e", 'ē': "e", 'ě': "e",
	'ì': "i", 'í': "i", 'î': "i", 'ï': "i", 'ī': "i",
	'ł': "l",
	'ñ': "n", 'ń': "n", 'ň': "n",
	'ò': "o", 'ó': "o", 'ô': "o", 'õ': "o", 'ö': "o", 'ø': "o", 'ō': "o",
	'œ': "oe",
	'ř': "r",
	'ś': "s", 'š': "s", 'ş': "s", 'ß': "ss",
	'ť': "t", 'þ': "th",
	'ù': "u", 'ú': "u", 'û': "u", 'ü': "u", 'ū': "u",
	'ý': "y", 'ÿ': "y",
	'ź': "z", 'ż': "z", 'ž': "z",
}