Re: Word-forward/word-backward as in bash / emacs

[Konstantine Rybnikov <k-bx@k-bx.com>]

06/04/2012 12:36 AM, Peter Stephenson написал:
> On Sun, 03 Jun 2012 23:28:03 +0300
> Konstantine Rybnikov<k-bx@k-bx.com>  wrote:
>> Only answer I got is to do "select-word-style bash", which is absolutely
>> not the same as bash (I also provided an example there that explains "why").
> So, with "select-word-style bash" that forward-word thing is the only
> difference, right?  It seemed to be when I tried your example.
>
> I actually noted forward-word was a bit funny when I wrote the function
> forward-word-match that implements it when you uses select-word-style.
> See if this does the trick --- it patches the function forward-word-match
> which will be installed in your $fpath.
>
> I've made the old behaviour a separate (non-default) option for the sake
> of this, but if this is what's needed we probably need to think a bit
> more about when it should apply.  The original behaviour works better
> with zsh's normal set up because then more characters are considered
> part of a word, so skipping non-word-characters at the start is less
> likely to have side effects.
>
> One possibility would be to reverse the sense of the option (to
> skip-whitespace-first), but to set it to "true" from select word-style
> if it's not explicitly set and the new style is "bash".  That keeps
> compatibility in both cases.  However, there are other ways of doing it.
>
> (I also standardised the indentation while I was there, so there's more
> changed than is really necessary.)
>
> Index: Functions/Zle/forward-word-match
> ===================================================================
> RCS file: /cvsroot/zsh/zsh/Functions/Zle/forward-word-match,v
> retrieving revision 1.3
> diff -p -u -r1.3 forward-word-match
> --- Functions/Zle/forward-word-match	28 Jul 2010 13:33:59 -0000	1.3
> +++ Functions/Zle/forward-word-match	3 Jun 2012 21:22:36 -0000
> @@ -8,32 +8,35 @@ local -a matched_words
>   integer count=${NUMERIC:-1}
>
>   if (( count<  0 )); then
> -    (( NUMERIC = -count ))
> -    zle ${WIDGET/forward/backward}
> -    return
> +  (( NUMERIC = -count ))
> +  zle ${WIDGET/forward/backward}
> +  return
>   fi
>
>   while (( count-- )); do
> -
> -    match-words-by-style
> -
> +  match-words-by-style
> +
> +  if zstyle -t $curcontext skip-whitespace-last; then
>       # For some reason forward-word doesn't work like the other word
>       # commands; it skips whitespace only after any matched word
>       # characters.
> -
>       if [[ -n $matched_words[4] ]]; then
> -        # just skip the whitespace
> -	word=$matched_words[4]
> -    else
> -        # skip the word and trailing whitespace
> -	word=$matched_words[5]$matched_words[6]
> -    fi
> -
> -    if [[ -n $word ]]; then
> -	(( CURSOR += ${#word} ))
> +      # just skip the whitespace
> +      word=$matched_words[4]
>       else
> -	return 1
> +      # skip the word and trailing whitespace
> +      word=$matched_words[5]$matched_words[6]
>       fi
> +  else
> +    # more standard behaviour: skip leading whitespace and the word.
> +    word=$matched_words[4]$matched_words[5]
> +  fi
> +
> +  if [[ -n $word ]]; then
> +    (( CURSOR += ${#word} ))
> +  else
> +    return 1
> +  fi
>   done
>
>   return 0
Ok, it looks like Moritz Bunkus updated he's answer with working 
implementation of forward-word-match. You're right, the only difference 
if forward-word things.

Just tried your patch -- yes, looks like it works as expected. Thanks!