Glenn Fowler schrieb:
On Sun, 28 Oct 2007 12:17:39 +0100 Bernd Eggink wrote:
Both $'\xe4' and $'\303\244' render as a-umlaut with LANG=de_DE.UTF-8.
Only $'\303\244' does so with LANG=C.
I'm not quite convinced that this isn't a ksh issue...
Thanks anyway!
Confused,
Bernd
confused here too
you and I aren't getting off that easy
part of the problem is that I'm usually in LANG=C
and am oblivious to many of the locale subtleties
so the following analysis could be off base
please jump in and correct any errors in the logic
I do know this much about utf-8 encoding
the leftmost 1-bits in each utf-8 byte specify the number
of current and remaining bytes to make up the encoded character
Actually it's a bit more complicated, see here:
http://www.cl.cam.ac.uk/~mgk25/unicode.html#utf-8
is there a tty setting that says "if its not utf-8, try 8-bit ascii"?
Not that I know of, but the tty appears to behave so.
As to the original problem: I tried some other sequences, such as
$ a=(name='1ä') # 1 a-umlaut
$ print ${a.name}
1ä # correct: 1 a-umlaut
$ print $a
( name=$'1\344' ) # Huh?
$ print "$a"
(
name=$'1\344'
)
0344 or 0xe4 is the unicode value of a-umlaut, the correct utf-8 encoding
would be 0xc3a4. The results suggest that, under certain circumstances, the
output encoding of structured values is incorrect or simply missing.
Regards,
Bernd
--
Bernd Eggink
[EMAIL PROTECTED]
http://sudrala.de
_______________________________________________
ast-developers mailing list
[email protected]
https://mailman.research.att.com/mailman/listinfo/ast-developers
