Unicode Explained

size storage units for all characters. The encodings used in practice tend to be more

complicated. Methods for performing code point order sorting on UTF-8 and other

Unicode encodings are described in the Unicode standard in section 5.17 “Binary Or-

der.” It discusses the even more technical orders based on numeric ordering of code

units (such as octets that constitute UTF-8 encoded data) rather than code numbers.

Problems of legacy software

In simple programming tasks, comparisons of character and string data are sometimes

based on comparisons of code points. This applies basically to basic Latin letters in

contexts where it can be assumed (or it just is assumed) that we need not deal with any

other letters and that the case of letters is fixed (e.g., due to previous case folding). This

explains code like if((ch >= 'A') & (ch <= 'Z')) for testing whether the value of ch

is an (uppercase) letter. Such code can be efficient, but nowadays it is usually better to

use library subroutines (e.g., if(isletter(ch))), making the code more readable and

more portable without sacrificing efficiency. We will discuss such methods in Chapter

11.

As a user of programs, you may encounter sorted data and sorting routines that apply

to simple code point order. For example, if you use a tool for automatic generation of

an index for a publication, you might notice that the index will be sorted that way. If

the entries ...