# [PATCH 4/8] bash: Add space after unique command name is completed.

3 messages from 2007-02-04 to 2007-02-04. Participants: Shawn O. Pearce, Junio C Hamano.
Thread: https://gitlist.dev/t/6663

## Shawn O. Pearce, 2007-02-04 07:38

Subject: [PATCH 4/8] bash: Add space after unique command name is completed.
Message-ID: <20070204073827.GD17603@spearce.org>
URL: https://gitlist.dev/e/20070204073827.GD17603%40spearce.org

```
Because we use the nospace option for our completion function for
the main 'git' wrapper bash won't automatically add a space after a
unique completion has been made by the user.  This has been pointed
out in the past by Linus Torvalds as an undesired behavior.  I agree.

We have to use the nospace option to ensure path completion for
a command such as `git show` works properly, but that breaks the
common case of getting the space for a unique completion.  So now we
set IFS=$'\n' (linefeed) and add a trailing space to every possible
completion option.  This causes bash to insert the space when the
completion is unique.

Signed-off-by: Shawn O. Pearce <spearce@spearce.org>
---
 contrib/completion/git-completion.bash |   26 ++++++++++++++++++++------
 1 files changed, 20 insertions(+), 6 deletions(-)

diff --git a/contrib/completion/git-completion.bash b/contrib/completion/git-completion.bash
index 93f2af5..1cf576e 100755
--- a/contrib/completion/git-completion.bash
+++ b/contrib/completion/git-completion.bash
@@ -61,6 +61,20 @@ __git_ps1 ()
 	fi
 }
 
+__gitcomp ()
+{
+	local all c s=$'\n' IFS=' '$'\t'$'\n'
+	for c in $1; do
+		case "$c" in
+		--*=*) all="$all$c$s" ;;
+		*)     all="$all$c $s" ;;
+		esac
+	done
+	IFS=$s
+	COMPREPLY=($(compgen -W "$all" -- "${COMP_WORDS[COMP_CWORD]}"))
+	return
+}
+
 __git_heads ()
 {
 	local cmd i is_hash=y dir="$(__gitdir "$1")"
@@ -787,12 +801,12 @@ _git ()
 	done
 
 	if [ $c -eq $COMP_CWORD -a -z "$command" ]; then
-		COMPREPLY=($(compgen -W "
-			--git-dir= --version --exec-path
-			$(__git_commands)
-			$(__git_aliases)
-			" -- "${COMP_WORDS[COMP_CWORD]}"))
-		return;
+		case "${COMP_WORDS[COMP_CWORD]}" in
+		--*=*) COMPREPLY=() ;;
+		--*)   __gitcomp "--git-dir= --bare --version --exec-path" ;;
+		*)     __gitcomp "$(__git_commands) $(__git_aliases)" ;;
+		esac
+		return
 	fi
 
 	local expansion=$(__git_aliased_command "$command")
-- 
1.5.0.rc3.22.g5057

```

## Junio C Hamano, 2007-02-04 08:08

Subject: Re: [PATCH 4/8] bash: Add space after unique command name is completed.
Message-ID: <7vr6t6xs3d.fsf@assigned-by-dhcp.cox.net>
URL: https://gitlist.dev/e/7vr6t6xs3d.fsf%40assigned-by-dhcp.cox.net
In-Reply-To: <20070204073827.GD17603@spearce.org>

```
"Shawn O. Pearce" <spearce@spearce.org> writes:

> +__gitcomp ()
> +{
> +	local all c s=$'\n' IFS=' '$'\t'$'\n'
> +	for c in $1; do
> +		case "$c" in
> +		--*=*) all="$all$c$s" ;;
> +		*)     all="$all$c $s" ;;
> +		esac
> +	done
> +	IFS=$s
> +	COMPREPLY=($(compgen -W "$all" -- "${COMP_WORDS[COMP_CWORD]}"))
> +	return
> +}
> +

I do not understand what is going on here.  Care to explain?

 (1) "for c in $1": because $1 is not dquoted, it is split into
     word with IFS -- which IFS applies to this splitting?  The
     local one you hand-define to the typical " \t\n" value?

 (2) when this function returns, what IFS value does the caller
     would have?  Is the "local" used to avoid screwing up the
     caller?

```

## Shawn O. Pearce, 2007-02-04 08:13

Subject: Re: [PATCH 4/8] bash: Add space after unique command name is completed.
Message-ID: <20070204081301.GE16766@spearce.org>
URL: https://gitlist.dev/e/20070204081301.GE16766%40spearce.org
In-Reply-To: <7vr6t6xs3d.fsf@assigned-by-dhcp.cox.net>

```
Junio C Hamano <junkio@cox.net> wrote:
> "Shawn O. Pearce" <spearce@spearce.org> writes:
> 
> > +__gitcomp ()
> > +{
> > +	local all c s=$'\n' IFS=' '$'\t'$'\n'
> > +	for c in $1; do
> > +		case "$c" in
> > +		--*=*) all="$all$c$s" ;;
> > +		*)     all="$all$c $s" ;;
> > +		esac
> > +	done
> > +	IFS=$s
> > +	COMPREPLY=($(compgen -W "$all" -- "${COMP_WORDS[COMP_CWORD]}"))
> > +	return
> > +}
> > +
> 
> I do not understand what is going on here.  Care to explain?
> 
>  (1) "for c in $1": because $1 is not dquoted, it is split into
>      word with IFS -- which IFS applies to this splitting?  The
>      local one you hand-define to the typical " \t\n" value?

Correct.
 
>  (2) when this function returns, what IFS value does the caller
>      would have?  Is the "local" used to avoid screwing up the
>      caller?
> 

Correct.


Basically at the time I'm forming 'all' I want $1 split based on
normal splitting rules, as I'm using space, tag and LF within the
script in some places for the first parameter to __gitcomp.

But when I run compgen I need IFS to be just LF ($s) so that $all
is split only on LF and not on spaces, as spaces were added to $all
after each unique option.  This way the trailing space is part of
the unique option, and bash will tack in the trailing space for me,
despite having -o nospace enabled.

-- 
Shawn.

```