CJKV Information Processing, 2nd Edition

212

|

Chapter 4: Encoding Methods

NCRs are useful for environments that ultimately produce XML or HTML les. e char-

acters represented by NCRs are explicitly identied as Unicode characters, yet benet

from the strength and persistence made possible by an ASCII-based notation. NCRs are

useful to represent hard-to-input characters, or characters whose underlying representa-

tion may be ambiguous when displayed in binary form. Take for instances the two Uni-

code characters “” (41) and “” (21). Both represent the same uppercase Latin

letter; however, the former is proportional, but sometimes implemented using half-width

metrics, and the latter is explicitly full-width. When it is critical to unambiguously use one

character over another, in the context of web documents, using

Ａ to represent the

full-width uppercase Latin letter “A” has great value. Interestingly, dierentiating Latin “A”

(41), Cyrillic “A” (41), and Greek “A” (391) is an even greater feat, and NCRs

become especially useful when dierentiating such characters becomes critical.

NCRs do, however, require up to 10 bytes for representation, so their advantages in the

contexts in which they can be used come at the expense of increased string size.

ere are other NCR-like notations, oen referred to as escaping, whose use should be

pointed out or described, such the

\uXXXX notation as used ...