perl-unicode

Re: Unicode aware module

1999-06-14 05:13:01
On Sun, 13 Jun 1999 at 15:11:49 -0400, Ilya Zakharevich wrote:

We are discussing Perl here, not C.  In Perl, there is no access to
memory storage, so at best a usage of such terms is misleading.  Perl
works in much higher terms, like "number" or "string".

Not counting pack( 'p', $string ) that is, of course.

But, more relevantly, Perl does define operators that treat strings as
octet-sequences, viz. the bitwise-boolean family, and 'vec' does a
pretty convincing imitation, too.

As a humble Perl user, I'd be happy with "length", "substr", "index"
etc. doing their stuff on these funny "string" things - but they better
be consistent! - but "vec" and "pack('a'..)" are widely used by people
who assume that byte == character == octet.

(Aside: has perl ever run on machines with other than 8-bit bytes? There
aren't many of them around. Perl grew up in the shortish era when
8-bit characters were almost universal.)

My apologies if this is stamping old ground, but don't you need an extra
type internally to do things 'right', viz. "Octet Value" to go with the
String and numerical values? It's either that or *very* slow "length"
calculations.

Ian
-- 
    I am confident this explanation will dispell any feelings
    of certainty that may have been troubling you.
     - BWHOLMES(_at_)SJSUVM1(_dot_)sjsu(_dot_)edu (Cabbage) in 
<9601221753(_dot_)AA27669(_at_)spock>

<Prev in Thread] Current Thread [Next in Thread>