Re: OT: Gawk match() and numbers in scientific notation

On May 8, 6:11=A0pm, pk <p...@[EMAIL PROTECTED]
> wrote:
> On Thursday 8 May 2008 17:46, Hermann Peifer wrote:
>
> > My /usr/share/i18n/locales/en_GB has also inside it:
>
> > LC_CTYPE
> > copy "i18n"
>
> Yes, mine too. I just extrapolated a sample section.
>
> > i18n in return defines the character classes: upper, lower, etc. Each
> > one with a list of =A0Unicode version 5.0.0 code points, that are
> > included in a given character class.
>
> However, that file is included by many other files in that directory
> (notably, not by POSIX).
>
> $ grep -l 'copy "i18n"' /usr/share/i18n/locales/* | wc -l
> 99
>
> Does that mean that all these 99 locales have all the same characters in
> their upper, alpha, etc. classes?
>

I would think so. All these 99 locales are Unicode-aware, so why
should they differ in their understanding of what is an upper case or
lower case, character?

This is defined in http://www.unicode.org/Public/UNIDATA/UnicodeData.txt
(and related files in this directory).

Code point U+0041 is e.g. defined as:
0041;LATIN CAPITAL LETTER A;Lu;0;L;;;;;N;;;;0061;

Property value "Lu" stands for upper case letter and all Unicode-aware
locales will observe these properties (e.g. via the character class
definitions in file i18n).


> And, even if it is so, that helps only for POSIX character classes; how
do=
es
> one answer the question "what characters match, eg, '[A-Z]' under a
given
> locale?". I think the answer to this lies in the understanding of
> LC_COLLATE.
>

I guess you are right but I don't have a clear idea myself. I will
also have to read more in the documentation.

Hermann