Description:
------------
`mb_ereg`, `mb_ereg_search_regs` & `mb_ereg_search_getregs` return only numbered capturing groups, but they should return both numbered and named capturing groups.

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

mb_ereg('(?<wsp>\s*)(?<word>\w+)', '  中国', $m);
var_dump($m);

mb_ereg_search_init('  中国？');
mb_ereg_search('(?<wsp>\s*)(?<word>\w+)(?<punct>[？！])');
var_dump(mb_ereg_search_getregs());

Expected result:
----------------
array(5) {
  [0]=>
  string(8) "  中国"
  [1]=>
  string(2) "  "
  [2]=>
  string(6) "中国"
  ["wsp"]=>
  string(2) "  "
  ["word"]=>
  string(6) "中国"
}

array(7) {
  [0]=>
  string(11) "  中国？"
  [1]=>
  string(2) "  "
  [2]=>
  string(6) "中国"
  [3]=>
  string(3) "？"
  ["punct"]=>
  string(3) "？"
  ["wsp"]=>
  string(2) "  "
  ["word"]=>
  string(6) "中国"
}

Actual result:
--------------
array(3) {
  [0]=>
  string(8) "  中国"
  [1]=>
  string(2) "  "
  [2]=>
  string(6) "中国"
}

array(4) {
  [0]=>
  string(11) "  中国？"
  [1]=>
  string(2) "  "
  [2]=>
  string(6) "中国"
  [3]=>
  string(3) "？"
}