VDISABLEVAL should be -1 for HP-UX?

VDISABLEVAL should be -1 for HP-UX?

[Rodger V. Anderson]

Some time ago (a year and a half) I asked about a problem I was having
with the "set-mark-command" being ignored when it was ^@ (ascii nul).
I think I have figured out what the problem is.

The routine setterm() in zle_main.c uses VDISABLEVAL, as defined in
system.h with a value of 0, to disable certain keys.  Unfortunately,
HP-UX appears to treat a value of 0 as valid character, namely ^@
(ascii nul), instead of as undefined.  I've recompiled zsh with
VDISABLEVAL set to -1 and the set-mark-command now works.

Any comments?  I can provide more details if anyone is interested.

Rodger Anderson

Re: VDISABLEVAL should be -1 for HP-UX?

[Zoltan Hidvegi]

> Some time ago (a year and a half) I asked about a problem I was having
> with the "set-mark-command" being ignored when it was ^@ (ascii nul).
> I think I have figured out what the problem is.
> 
> The routine setterm() in zle_main.c uses VDISABLEVAL, as defined in
> system.h with a value of 0, to disable certain keys.  Unfortunately,
> HP-UX appears to treat a value of 0 as valid character, namely ^@
> (ascii nul), instead of as undefined.  I've recompiled zsh with
> VDISABLEVAL set to -1 and the set-mark-command now works.

This patch should fix the problem.

Zoltan


*** Src/system.h	1997/05/06 05:56:36	3.1.2.3
--- Src/system.h	1997/04/28 05:44:20
***************
*** 257,263 ****
  # else
  #  include <termios.h>
  # endif
! # define VDISABLEVAL 0
  # define HAS_TIO 1
  #else    /* not TERMIOS */
  # ifdef HAVE_TERMIO_H
--- 257,267 ----
  # else
  #  include <termios.h>
  # endif
! # ifdef _POSIX_VDISABLE
! #  define VDISABLEVAL _POSIX_VDISABLE
! # else
! #  define VDISABLEVAL 0
! # endif
  # define HAS_TIO 1
  #else    /* not TERMIOS */
  # ifdef HAVE_TERMIO_H