Description:
------------
mb_ereg_search_getpos() seems to count German umlauts twice. See test script.

Test script:
---------------
mb_regex_encoding('UTF-8');

$test1 = "ab";
$test2 = "ääb";
$test3 = "äääb";

mb_ereg_search_init($test1, "b");
mb_ereg_search();
echo "Test1: " . mb_ereg_search_getpos(); // Result: 2

mb_ereg_search_init($test2, "b");
mb_ereg_search();
echo "Test2: " . mb_ereg_search_getpos(); // Result: 5 (should be "3")

mb_ereg_search_init($test3, "b");
mb_ereg_search();
echo "Test3: " . mb_ereg_search_getpos(); // Result: 7 (should be "4")