HTML export should leave IRIs as is

["Vladimir Alexiev" <vladimir@sirma.bg>]

I have this in a file:

- [[https://chrome.google.com/webstore/detail/copy-url/mkhnbhdofgaendegcgbmndipmijhbili][CopyURL]] Chrome Plugin
  - Many URLs in dbpedia are International Resource Identifiers (IRIs)
  - When you copy from the browser's address bar, an IRI is URL-encoded
  - Unreadable ugliness like http://bg.dbpedia.org/resource/%D0%9A%D0%BE%D0%BD%D1%81%D1%82%D0%B0%D0%BD%D1%82%D0%B8%D0%BD%D0%B0_%D0%96%D0%B8%D0%B2%D0%BE%D0%B2%D0%B0
  - Instead, this plugin allows you to copy & paste IRIs like http://bg.dbpedia.org/resource/Константина_Живова

With default settings, the HTML export made both URLs the same unreadable ugliness.

I see no good reason for ox-html to escape IRIs, I think it should emit them as IRIs.
- I expected (setq org-html-use-unicode-chars t) to fix this, but for some reason (mm-url-decode-entities) doesn't do it
- This fixed the problem: (org-url-hexify-p nil), but I'm afraid it may break something else.

Furthermore, this fix converted BOTH URLs to an IRI. So to illustrate the point, I had to change my org file to:

  - Unreadable ugliness like
    : http://bg.dbpedia.org/resource/%D0%9A%D0%BE%D0%BD%D1%81%D1%82%D0%B0%D0%BD%D1%82%D0%B8%D0%BD%D0%B0_%D0%96%D0%B8%D0%B2%D0%BE%D0%B2%D0%B0