git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Junio C Hamano <gitster@pobox.com>
To: Siddharth Asthana <siddharthasthana31@gmail.com>
Cc: git@vger.kernel.org, "Ævar Arnfjörð Bjarmason" <avarab@gmail.com>,
	"Derrick Stolee" <derrickstolee@github.com>,
	"Elijah Newren" <newren@gmail.com>,
	"Lars Schneider" <larsxschneider@gmail.com>
Subject: Re: [GSoC] [PATCH] t1011: replace test -f with test_path_is_file
Date: Mon, 11 Apr 2022 12:09:28 -0700	[thread overview]
Message-ID: <xmqq1qy3igif.fsf@gitster.g> (raw)
In-Reply-To: <20220409114458.23435-1-siddharthasthana31@gmail.com> (Siddharth Asthana's message of "Sat, 9 Apr 2022 17:14:56 +0530")

Siddharth Asthana <siddharthasthana31@gmail.com> writes:

> Use test_path_is_file() instead of 'test -f' for better debugging
> information.
> ---

missing Sign-off.

>  	test_cmp expected.swt result &&
> -	! test -f init.t &&
> -	! test -f sub/added
> +	! test_path_is_file init.t &&
> +	! test_path_is_file sub/added
>  '

Given the definition of the helper function, i.e.

        test_path_is_file () {
                test "$#" -ne 1 && BUG "1 param"
                if ! test -f "$1"
                then
                        echo "File $1 doesn't exist"
                        false
                fi
        }

the new test will _complain_ "init.t doesn't exist" when we have
successfully run the test, while it will be _silent_ when init.t
that _should_ not exist is there.

Which is the complete opposite of the spirit of why we want to use
the helper when we expect the path "$1" to exist, i.e. loudly fail
when our expectation is _not_ met.

$ git grep '! test_path_is' t/

shows that we already have such a misuse of test_path_is_dir in one
place, but luckily we do not have any for test_path_is_file or other
similar helpers.  test_path_is_hidden is sort-of OK as that is not
about verbosity.

In these two test, we do not expect init.t or sub/added to _exist_
at all.  It's not like we are happy if we see init.d exist as a
directory (which is not a file).  test_path_is_missing is probably
the right helper to use.

It is not very plausible that we'd want to assert that existence of
a path as a file the only bad condition (i.e. we are happy if the
path did not exist or it is a directory, symlink, or a socket), so I
think the simple 

	Never use '! test_path_is_file'; test_path_is_missing may be
	what you are looking for.

is a good enough rule.

If not, we could allow the caller to write such a convoluted "only
existence of a path as a file is unacceptable and everything else is
good" assertion as

    test_path_is_file ! init.d

with something like

        test_path_is_file () {
		expecting_file=true
		if test "$1" = "!"
		then
			expecting_file=false
			shift
		fi
                test "$#" -ne 1 && BUG "1 param"
                if test -f "$1"
                then
                	$expecting_file || echo "File $1 exists"
                        $expecting_file
		else
			$expecting_file && echo "File $1 doesn't exist"
                        ! $expecting_file
                fi
        }

but I do not think we want to go that way.

  reply	other threads:[~2022-04-11 19:09 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-04-09 11:44 [GSoC] [PATCH] t1011: replace test -f with test_path_is_file Siddharth Asthana
2022-04-11 19:09 ` Junio C Hamano [this message]
2022-04-12 20:21   ` Siddharth Asthana
2022-04-12 20:37   ` [GSoC] [PATCH v2] " Siddharth Asthana
2022-04-14  8:19     ` Christian Couder
2022-04-14 16:42       ` Junio C Hamano
2022-04-16 13:55         ` Siddharth Asthana
2022-04-16 13:59           ` [PATCH v3] t1011: replace test -f with test_path_is* helpers Siddharth Asthana

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=xmqq1qy3igif.fsf@gitster.g \
    --to=gitster@pobox.com \
    --cc=avarab@gmail.com \
    --cc=derrickstolee@github.com \
    --cc=git@vger.kernel.org \
    --cc=larsxschneider@gmail.com \
    --cc=newren@gmail.com \
    --cc=siddharthasthana31@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).