Permalinks in Double Byte languages

URIs use an 82 character subset of US ASCII as defined in RFC 2396. There are schemes for encoding i18n URIs. To assure uniform encoding, UTF-8 is dictated. This isn’t too difficult for someone who is operating their blog in UTF-8. Simply replace all non RFC 2396 characters with their UTF-8 hex value. If the blog is not operating in UTF-8, we get into the mess of juggling encodings. This is why we added code to CVS to simply use the post id if sanitize_title returns an empty string. If someone wants to write a sanitize_title plugin that performs i18n URI encoding, go for it. Take a look at seems_utf8() and remove_accents() for some hints. If the plugin is general purpose enough to handle all scenarios, we could incorporate it into WordPress. Anyone want to have a go?

Yes, in the nightly. We do this for posts and categories. I still need to fix authors. We might make the category and author “nice names” editable in the manner of the post slug.