Question

According to this article, the right way to support emoji characters from iOS4 through iOS6 is to use utf8mb4 encoding. Does App Engine’s datastore support utf8mb4? Currently I use the Java version of App Engine as the back end in my iOS app.

Answer 1

Firstly, the article you linked to talks about the writer’s experience with developing apps for iOS using mysql as the database. In mysql, the encoding “utf8” represents only characters that have 1, 2 or 3 bytes length.

However, utf8 by definition supports 1, 2, 3, or 4 byte long characters. So in the context of mysql encodings and collations, utf8 means characters only upto 3 bytes. For 4 byte long characters, mysql has devised the utf8mb4 encoding.

However, this applies only to the mysql context. Outside of the mysql world, utf8 automatically means support for 1,2,3, or 4 byte characters. See the UTF-8 wikipedia page, it talks about this in the introductory paragraphs:

UTF-8 encodes each of the 1,112,064[7] code points in the Unicode
character set using one to four 8-bit bytes (termed “octets” in the
Unicode Standard).

So when a platform or software application claims to support UTF-8, it implies support for all the 1,112,064 code points that can consume 1,2,3, or 4 bytes per character.

The mysql case is an anomaly where the utf8 encoding wasn’t really 100% utf8 but only some 6% of utf-8 (but most people wouldn’t notice that unless they start using emojis or other characters).

So my suggestion is that it is safe to assume that google app engine supports 4 byte characters when they accept utf-8 as a valid encoding supported by them.