Re: [gawk] RE bug??

[Paul Jarc]

Stephen Davies <address@hidden> wrote:
> It also seems strange to me that Bob's en_US should treat lower case 
> letters as upper case. Doesn't sound right to me.

It doesn't treat lowercase letters as uppercase.  It just alternates
between lowercase and uppercase letters for sorting purposes.  [A-Z]
actually means "A, Z, and all the characters (not necessarily just
uppercase letters) that come between them in the current locale's sort
order".  In the C locale, that's equivalent to [[:upper:]], since the
uppercase letters are consecutive in the sort order.  But other
locales may sort characters in an order like AaBbCc...Zz or
aAbBcC...zZ.  Uppercase letters are still distinct from lowercase, but
they are no longer consecutive in the sort order, and so a
sort-order-based range is no longer useful as a way to specify a
character class of uppercase or lowercase letters.


paul