I would like a Greek word (Ἀπὸ) to render on a page like it rendered right here. If I copy Ἀπὸ directly to a page via the Wordpress interface, it shows just fine in both the interface and on the actual page. However when I use some PHP to read it out of a MySQL table and add it to a page, it renders ??? instead of Ἀπὸ.
Before you think it is messed up in the database, it renders as expected in Sequel Pro. However I did notice something interesting in its interface. You can select "View Using Encoding..." If I select "UTF-8 Unicode BMP (utf8)" or "UTF-8 Full Unicode (utf8mb4)", it works. If I select "UTF-8 Unicode via Latin 1", I get ???.
The database encoding is: UTF-8 Unicode (utf8)
The database collation is: utf8_unicode_ci
The table reports the same encoding and collation.
If I view the source of the page that gives me "???" I see:
<meta charset='utf-8'>
What am I missing?