The manner in which TERR parses character types aids in internationlization.

The TERR parser can process text containing Unicode characters. All S language keywords (function, for, and so on) are ASCII. Numbers are parsed only when written with ASCII digits using the US number format (such as "1234.56").

All non-ASCII Unicode characters are treated as possible name characters, except for space characters. Thus, it is possible to have a data.frame column name, a factor level, or a variable name containing Japanese characters.