Auto generated ID for h1-h6 should support multi-language

[mytharcher]

This issue is additional about #110's duplicated ID problem.

It still happen in case that the title contains other language characters like Chinese words. For example:

## abc之一
## abc之二

than the compiled html output this:

<h2 id="abc">abc之一</h2>
<h2 id="abc">abc之二</h2>

It ignored the characters unless [a-z]. In neither maruku nor rdiscount it can work right.

And I highly recommend that it should output all the title string as same as anyone wrote. Unless these cases:

1. If the title has been used once, than add a unique random suffix;
2. If the title contains space, than use hyphen "-" instead;

I know that it may be the markdown engine's affair, but I saw the maruku's bug list hadn't update since 2011-04-14 02:23. And this issue happen in my using Jekyll, so I post here for asking help.

Hope you can help, thank you!

[stereobooster]

Try MultiMarkdown (#379) or Redcarpet (#450, #487)

[mytharcher]

Thank you! I also found Kramdown works right.