cyrillic translit

I tried to find a generic, standard solution to the transliteration problem, when generating urls from titles. Let me stress it's quite important, if we don't want Drupal to be limited to latin alphabets.

I eventually wrote stg around to the iconv function:

$separator = '_';
$pattern = '/[^a-zA-Z0-9 ]+/';

$txt = iconv("UTF-8", "ASCII//TRANSLIT", $txt);
$txt = strtolower($txt);
$txt = trim(preg_replace($pattern, '', $txt));
$txt = preg_replace('/ +/', $separator, $txt);

I don't know how well this works exactly. In fact, where my site is hosted, iconv is not supported :(

Can you have a look at how the problem is solved in textpattern?
They have an external translit map which, though not complete, can be extended without touching the code. It already includes rules for accented chars, cyrillic and greek.

http://dev.textpattern.com/browser/development/4.0/textpattern/lib/txplib_misc.php?rev=840
http://dev.textpattern.com/browser/development/4.0/textpattern/lib/i18n-ascii.txt

The most interesting function is dumbDown. It's very simple, and could be merged with the existing code in few minutes.
(Textpattern is free, open source software, distributed under the GNU General Public License.)

Thank you! Ciao. Mic

As I don't know about a windows version of diff which supports unicode, I'm posting here my 'patch' to transliterate urls (it works for nodes, users, and everything).

The following lines should be inserted soon after the definition of $translations (in cleanstring - pathauto.module). Also, you should add this file to the pathauto folder.

  static $i18n_loaded = false;
  if (! $i18n_loaded) {
    $path = drupal_get_path('module', 'pathauto');
    if (is_file($path. '/i18n-ascii.txt')) {
      $i18n = parse_ini_file($path. '/i18n-ascii.txt');
      $translations = array_merge($translations, $i18n);
    }
    $i18n_loaded = true;
  }