Unicode

Character Array and Name Storage

All character arrays and names (variable, function, and operator) are stored as one 16-bit word per character. This fixed length encoding is called UCS-2 and is a subset of a more general encoding called UTF-16. The latter is a variable length encoding using one or two 16-bit words per character.

UCS-2 represents characters in the range U+0000 through U+FFFF; UTF-16 represents characters in the range U+0000 through U+10FFFF. However, because of the way UTF-16 represents characters above U+FFFF, the range of code points for both encodings exclude U+D800 through U+DFFF (2,048 characters).

Thus, UCS-2 encodes 63,488 (=65,536 - 2,048) different characters, and UTF-16 encodes 1,112,064 (=1,114,112 - 2,048) different characters.

Alphabet for Names

The alphabet used for names consists of an initial character followed by one or more subsequent characters.

• An initial character is one of a though z, A through Z, delta (∆), or delta underbar (⍙).
• A subsequent character is a leading character, 0 through 9, overbar (¯), or underbar (_).

One other set of characters, the underbarred alphabet ( through ), may be pasted into a session or function editor window. There is no way to enter these characters directly from the keyboard. Depending on a User Option setting, when these characters are pasted into a session or function editor window, they are treated as themselves or are mapped to the lowercase alphabet. When used in a name, they are always equivalent to the corresponding lowercase letter, although they display as themselves. Because of this latter translation, they may be used as either a leading or subsequent character in a name. Thus the names lh and alpha display differently, but they both refer to the same object; a value assigned to one is reflected in the other.

See also: Quad AV = niladic system function ⎕AV - Atomic Vector page.