get_page_uri returns invalid URLs for multibyte page slug

[lilyfan]

If a page has a multibyte page slug (like %E4%BE%8B ; means "example" in japanese), get_page_uri() returns invalid URL as " http://example.com/例/". It is not compliant with RFC 1748, because a UTF-8 character is inlcuded in the path. The reason is that get_page_uri() does "urldecode($page->post_name)". Thus "%E4%BE%8B" is decoded to "例" (a UTF-8 character).

The solution is to avoid using urldecode().

The urldecode calls were introduced long ago in [1841] (1.3-alpha-4) when for some reason page names were double encoded.

The patch is attached.

(In [8096]) Don't decode slug in get_page_uri(). Props nbachiyski. fixes #6723