Re: [PATCH 1/2] Add Git::config_path()

From: Jakub Narebski <jnareb@gmail.com>
Date: Oct 7, 2011, 21:44 UTC
Message-ID: <201110072344.46556.jnareb@gmail.com>
In-Reply-To: <7vk48gtrh9.fsf@alter.siamese.dyndns.org>

On Mon, 7 Oct 2011, Junio C Hamano wrote:

Show 17 quoted lines

> Jakub Narebski <jnareb@gmail.com> writes:
> 
> > diff --git a/t/t9700-perl-git.sh b/t/t9700-perl-git.sh
> > index 3787186..7558f0c 100755
> > --- a/t/t9700-perl-git.sh
> > +++ b/t/t9700-perl-git.sh
> > @@ -43,7 +43,9 @@ test_expect_success \
> >       git config --add test.booltrue true &&
> >       git config --add test.boolfalse no &&
> >       git config --add test.boolother other &&
> > -     git config --add test.int 2k
> > +     git config --add test.int 2k &&
> > +     git config --add test.path "~/foo" &&
> > +     git config --add test.pathexpanded "$HOME/foo"
> 
> Given that test-lib.sh sets up the $HOME away from unknown place to ensure
> repeatability of tests, I am not sure if this test would ever pass.

Well, it passes.

test-lib.sh sets $HOME to "$TRASH_DIRECTORY", but this value of $HOME is then later seen by "git config --path ..." run by Git::config_path in t9700/test.pl

  char *expand_user_path(const char *path)
  [...]
                if (username_len == 0) {
                        const char *home = getenv("HOME");
                        if (!home)
                                goto return_null;
                        strbuf_add(&user_path, home, strlen(home));
                } else {
  [...]

-- 
Jakub Narebski
Poland