Re: Echoing of 8-bit-characters broken after 4.3.2 (solved)

Re: Echoing of 8-bit-characters broken after 4.3.2 (solved)

[Wolfgang Hukriede]

Special thanks to Phil Pennock who made me see the light. Indeed it
suffices to

export LC_CTYPE=en_US.ISO8859-1

to get the old behavior back with 8-bit-chars echoed unchanged.
Tab-completion works also. This is for a shell with multibyte support
(--enable-multibyte) compiled in.

Bart, setting LANG would definitely be overkill, because I do not want
message language, time-format etc. to be affected. (FWIW, not even
a change to the collation sequence).

Now, since this is a silent change under the hood and I cannot believe
to be the only one to be hurt by it, may I suggest to place a hint
into the man page (or maybe at least the faq)? Something like (in
zshparams):

  LC_CTYPE <S>
       This variable determines the locale category for character
       handling functions. If you want 8-bit-characters echoed back
       and working tab completion on a system which does not use
       utf-8, consider to set the locale appropriately, e.g. export
       LC_CTYPE=en_US.ISO8859-1.

Many thanks to everyone!

Re: Echoing of 8-bit-characters broken after 4.3.2 (solved)

[Peter Stephenson]

On Sun, 1 Mar 2009 10:57:53 +0100
Wolfgang Hukriede <whukriede@googlemail.com> wrote:
> Now, since this is a silent change under the hood and I cannot believe
> to be the only one to be hurt by it, may I suggest to place a hint
> into the man page (or maybe at least the faq)?

Index: Doc/Zsh/params.yo
===================================================================
RCS file: /cvsroot/zsh/zsh/Doc/Zsh/params.yo,v
retrieving revision 1.51
diff -u -r1.51 params.yo
--- Doc/Zsh/params.yo	6 Feb 2009 09:59:38 -0000	1.51
+++ Doc/Zsh/params.yo	3 Mar 2009 20:44:56 -0000
@@ -906,7 +906,12 @@
 vindex(LC_CTYPE)
 item(tt(LC_CTYPE) <S>)(
 This variable determines the locale category for character handling
-functions.
+functions.  If the tt(MULTIBYTE) option is in effect this variable or
+tt(LANG) should contain a value that reflects the character set in
+use, even if it is a single-byte character set, unless only the
+7-bit subset (ASCII) is used.  For example, if the character set
+is ISO-8859-1, a suitable value might be tt(en_US.iso88591) (certain
+Linux distributions) or tt(en_US.ISO8859-1) (MacOS).
 )
 vindex(LC_MESSAGES)
 item(tt(LC_MESSAGES) <S>)(
Index: Doc/Zsh/roadmap.yo
===================================================================
RCS file: /cvsroot/zsh/zsh/Doc/Zsh/roadmap.yo,v
retrieving revision 1.11
diff -u -r1.11 roadmap.yo
--- Doc/Zsh/roadmap.yo	14 Apr 2008 12:53:36 -0000	1.11
+++ Doc/Zsh/roadmap.yo	3 Mar 2009 20:44:56 -0000
@@ -50,6 +50,17 @@
 There is some discussion of this in the shell FAQ,
 http://zsh.dotsrc.org/FAQ/ .  Note in particular that for combining
 characters to be handled the option tt(COMBINING_CHARS) needs to be set.
+Because the shell is now more sensitive to the definition of the
+character set, note that if you are upgrading from an older version of
+the shell you should ensure that the appropriate variable, either
+tt(LANG) (to affect all aspects of the shell's operation) or
+tt(LC_CTYPE) (to affect only the handling of character sets) is set to
+an appropriate value.  This is true even if you are using a
+single-byte character set including extensions of ASCII such as
+tt(ISO-8859-1) or tt(ISO-8859-15).  See the description of tt(LC_CTYPE)
+in
+ifnzman(noderef(Parameters))\
+ifzman(zmanref(zshparam)).
 
 subsect(Completion)
 

-- 
Peter Stephenson <p.w.stephenson@ntlworld.com>
Web page now at http://homepage.ntlworld.com/p.w.stephenson/

Re: Echoing of 8-bit-characters broken after 4.3.2 (solved)

[Wolfgang Hukriede]

Great, thank you, appreciated very much!

Best wishes, Wolfgang