PROMPT_SP and empty PROMPT_EOL_MARK

PROMPT_SP and empty PROMPT_EOL_MARK

[Mikael Magnusson]

The recent patches to PROMPT_SP broke my local hacks that removed the
ugly % sign, so I found out about PROMPT_EOL_MARK but why on earth
does it

              If not set or empty, the default  behavior
              is equivalent to the value `%B%S%#%s%b'.

What is the rationale for disallowing an empty indicator? If someone
wants the default they can just unset the parameter.

diff --git a/Src/utils.c b/Src/utils.c
index 8e557fd..8a2db5e 100644
--- a/Src/utils.c
+++ b/Src/utils.c
@@ -1285,7 +1285,7 @@ preprompt(void)
        char *eolmark = getsparam("PROMPT_EOL_MARK");
        char *str;
        int percents = opts[PROMPTPERCENT], w = 0;
-       if (!eolmark || !*eolmark)
+       if (!eolmark)
            eolmark = "%B%S%#%s%b";
        opts[PROMPTPERCENT] = 1;
        str = promptexpand(eolmark, 1, NULL, NULL, NULL);
diff --git a/Doc/Zsh/params.yo b/Doc/Zsh/params.yo
index 5d85f58..bec0027 100644
--- a/Doc/Zsh/params.yo
+++ b/Doc/Zsh/params.yo
@@ -1158,7 +1158,7 @@ item(tt(PROMPT_EOL_MARK))(
 When the tt(PROMPT_CR) and tt(PROMPT_SP) options are set, the
 tt(PROMPT_EOL_MARK) parameter can be used to customize how the end of
 partial lines are shown.  This parameter undergoes prompt expansion, with
-the tt(PROMPT_PERCENT) option set.  If not set or empty, the default
+the tt(PROMPT_PERCENT) option set.  If not set, the default
 behavior is equivalent to the value `tt(%B%S%#%s%b)'.
 )
 vindex(PS1)


-- 
Mikael Magnusson

Re: PROMPT_SP and empty PROMPT_EOL_MARK

[Bart Schaefer]

On Dec 5, 10:10pm, Mikael Magnusson wrote:
}
}               If not set or empty, the default  behavior
}               is equivalent to the value `%B%S%#%s%b'.
} 
} What is the rationale for disallowing an empty indicator?

Probably so that you can declare it as a local without having to
explicitly unset it afterwards.

You can get the effect you want with something like

PROMPT_EOL_MARK='%{%}'

or

PROMPT_EOL_MARK='%<<'

In fact because prompt expansion is applied to PROMPT_EOL_MARK *before*
the precmd functions are called, there are all sorts of interesting side
effects I hadn't previously considered ...

Re: PROMPT_SP and empty PROMPT_EOL_MARK

[Wayne Davison]

[-- Attachment #1: Type: text/plain, Size: 424 bytes --]

On Sun, Dec 5, 2010 at 1:10 PM, Mikael Magnusson <mikachu@gmail.com> wrote:

> What is the rationale for disallowing an empty indicator?


I had expected someone to set the char to a space if they didn't want a
visible indicator char.  However, I doubt that it would hurt to allow an
empty string to be used as well, since the C code seems fine at measuring a
zero-width size for an empty PROMPT_EOL_MARK string.

..wayne..

Re: PROMPT_SP and empty PROMPT_EOL_MARK

[Greg Klanderman]

> What is the rationale for disallowing an empty indicator?

If you don't want it, wouldn't you just turn off PROMPTSP?

greg

Re: PROMPT_SP and empty PROMPT_EOL_MARK

[Mikael Magnusson]

On 6 December 2010 21:35, Greg Klanderman <gak@klanderman.net> wrote:
>
>> What is the rationale for disallowing an empty indicator?
>
> If you don't want it, wouldn't you just turn off PROMPTSP?

Because that's not what promptsp does?

% echo -n hello
hello
% setopt nopromptsp
% echo -n hello
%
% setopt nopromptcr
% echo -n hello
hello%
(at this point the cursor is on the first l)

-- 
Mikael Magnusson

Re: PROMPT_SP and empty PROMPT_EOL_MARK

[Greg Klanderman]

>>>>> On December 6, 2010 Mikael Magnusson <mikachu@gmail.com> wrote:

> Because that's not what promptsp does?

Yes, of course, sorry about that.. your proposed change
makes sense, and I see pws has already committed it.
I hadn't really thought about it when I originally
submitted the patch that created PROMPT_EOL_MARK (pws
renamed the parameter if I recall correctly).

Greg