Greedy regex mangles successive tags

Two tags
Pre1<dme:x>foo</dme:x>Post1 Pre2<dme:x>bar</dme:x>Post2 , where dme:x will be replaced by §inner_content§ will result in Pre1§bar§Post2.

Obviously the two regexes for preparing and processing tags with close tags are too greedy.
In _dme_prepare_text I replaced
$regex_with_close_tag = "@<dme:(". implode('|', $tag_list) .')\s*((?:\s+\w+=([-+]?[0-9.]+|(["\']).+\4))*)\s*>(.*)</dme:\1>@isU';
by
$regex_with_close_tag = "@<dme:(". implode('|', $tag_list) .')\s*((?:\s+\w+=([-+]?[0-9.]+|(["\']).+\4))*)\s*>((?U).*)</dme:\1>@isU';
I'm not very familiar with PERL regexes but (.*?) did not work as I expected.

In _dme_process_text I did the same but to make it work fast I had to remove the nested tags logic (which I took no time to fully understand):
$regex = "@\[dme:(". implode('|', $tag_name_list) .')\s*((?:\s+\w+=([-+]?[0-9.]+|(["\']).+\4))*)\s*\]((?U).*)\[/dme:\1\]@isU';

I just took a very quick look into the code but I think that the solution with the tag nesting does not work anyway. A recursive approach would be better I think.

Regards,
Oliver