Expect localized space characters when string-processing user data, like U+3000 IDEOGRAPHIC SPACE

[chrisbobbe]

Some scripts use space characters that are different from the one I'm using to write this (U+0020 SPACE). For example, U+3000 IDEOGRAPHIC SPACE is used in Chinese, Japanese, Korean, and other languages.

• In autocomplete, we process queries and result data, like users' names, by splitting words with ' '.
• In compose, we insert ' ' after inserting generated syntax like for mentions.
• (Possibly other places.)

In each of these places, we should do what will seem most natural to the user.