From: Jeff King <peff@peff.net>
Date: Fri, 28 Sep 2012 19:30:08 GMT
Subject: Re: [PATCH 2/5] completion: fix args of run_completion() test helper
Message-ID: <20120928193008.GA3912@sigill.intra.peff.net>
In-Reply-To: <7vzk4aj6ik.fsf@alter.siamese.dyndns.org>

On Fri, Sep 28, 2012 at 12:23:47PM -0700, Junio C Hamano wrote:

> >> > @@ -57,7 +57,7 @@ run_completion ()
> >> >  test_completion ()
> >> >  {
> >> >  	test $# -gt 1 && echo "$2" > expected
> >> > -	run_completion "$@" &&
> >> > +	run_completion $1 &&
> >> >  	test_cmp expected out
> >> >  }
> >> 
> >> I can understand the other three hunks, but this one is fishy.
> >> Shouldn't "$1" be inside a pair of dq?  I.e.
> >> 
> >> 	+	run_completion "$1" &&
> >
> > No.  $1 holds all words on the command line.  If it was between a pair
> > of dq, then the whole command line would be passed to the completion
> > script as a single word.
> 
> And these "words" can be split at $IFS boundaries without any
> issues?  IOW, nobody would ever want to make words array in the
> run_completion function to ['git' 'foo bar' 'baz']?

It might be simpler to just convert test_completion into the
test_completion_long I added in my series; the latter takes the expected
output on stdin, leaving the actual arguments free to represent the real
command-line. E.g., your example would become:

  test_completion git "foo bar" baz <<-\EOF
  ... expected output ...
  EOF

-Peff