Three dashes in permalink = broken links

[maggiemasetti]

In the past Wordpress would generate a permalink for us. This was never an issue until we upgraded til Wordpress 3 - now the links that WP itself generated, which have three dashes in them, are broken.

For example:
The browser URL bar says this:
​http://astrophysics.gsfc.nasa.gov/outreach/podcast/wordpress/index.php/2011/05/09/blog-weekly-awesomeness-round-up---5911/

But when I cut/paste it, it comes across as this:
​http://astrophysics.gsfc.nasa.gov/outreach/podcast/wordpress/index.php/2011/05/09/blog-weekly-awesomeness-round-up-%E2%80%93-5911/

Either way, WP can apparently no longer parse three dashes and gives us an error that says "Not Found. Sorry, you are looking for something that isn't here." However, this entry is very much there - and we didn't create the URL, WP did. And it used to work just fine.

So now, we are left with the option of changing the existing link ourselves, but that will also break any links to this page.

We optimally need WP to fix it such such that the site can once again interpret permalinks that were created by WP with three dashes in them.

Currently we are using WP 3.2.1.

Here's my email address:
maggie.masetti@…

[SergeyBiryukov]

%e2%80%93 is –.

#10797 should fix this.

[nacin]

Hi Maggie,

I'm not sure the reason for the 404, as this should work without an issue. I'm not sure if it has to do with the fact that you're using ugly PATHINFO links (index.php in them). Those are lame, and sometimes buggy.

Unfortunately, the middle dash is an en dash. We don't sanitize these. Beginning in version 3.3 (currently in development), we should be. See #10797.

WordPress does remember old slugs for non-hierarchical post types. So if you edit the slug to remove the en dash (and thus the hyphens will collapse into one), it'll remember the old link. Assuming, of course, the link doesn't 404, as it is for you currently.

[SergeyBiryukov]

Replying to maggiemasetti:

So now, we are left with the option of changing the existing link ourselves, but that will also break any links to this page.

Since version 2.1, WordPress should properly redirect the old link to the new one (this was introduced in #3202).

[SergeyBiryukov]

Replying to nacin:

I'm not sure if it has to do with the fact that you're using ugly PATHINFO links (index.php in them).

This could be PATHINFO indeed, see my comment on #10249. The fix for get_page_by_path() is now committed, but the fix for parse_request() is needed too.