From: Austin Clements <amdragon@MIT.EDU>
To: Jani Nikula <jani@nikula.org>, notmuch@notmuchmail.org
Subject: Re: [PATCH v2 3/4] test: notmuch search --format=text0
In-Reply-To:
 <5c881edb2803e613e8726f71fe86772d372b38a0.1355064714.git.jani@nikula.org>
References: <cover.1355064714.git.jani@nikula.org>
	<5c881edb2803e613e8726f71fe86772d372b38a0.1355064714.git.jani@nikula.org>
User-Agent: Notmuch/0.14+184~gc26cc51 (http://notmuchmail.org) Emacs/23.4.1
	(i486-pc-linux-gnu)
Date: Sun, 16 Dec 2012 10:51:08 -0500
Message-ID: <87y5gyvvv7.fsf@awakening.csail.mit.edu>
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Transfer-Encoding: quoted-printable
Precedence: list

On Sun, 09 Dec 2012, Jani Nikula <jani@nikula.org> wrote:
> ---
>  test/text |   29 +++++++++++++++++++++++++++++
>  1 file changed, 29 insertions(+)
>
> diff --git a/test/text b/test/text
> index 428c89b..e003a66 100755
> --- a/test/text
> +++ b/test/text
> @@ -52,4 +52,33 @@ output=3D$(notmuch search --format=3Dtext "t=C3=ABxt-s=
earch-m=C3=A9ssage" | notmuch_search_s
>  test_expect_equal "$output" \
>  "thread:XXX   2000-01-01 [1/1] Notmuch Test Suite; text-search-utf8-body=
-s=C3=BCbj=C3=A9ct (inbox unread)"
>=20=20
> +add_email_corpus
> +
> +test_begin_subtest "Search message tags: text0"
> +cat <<EOF > EXPECTED.$test_count

Other tests use just OUTPUT and EXPECTED.  Why the $test_count?  Is
there a technical reason for it?

> +attachment inbox signed unread
> +EOF
> +notmuch search --format=3Dtext0 --output=3Dtags '*' | xargs -0 | notmuch=
_search_sanitize > OUTPUT.$test_count
> +test_expect_equal_file EXPECTED.$test_count OUTPUT.$test_count
> +
> +test_begin_subtest "Compare text vs. text0 for threads"
> +notmuch search --format=3Dtext --output=3Dthreads '*' | notmuch_search_s=
anitize > EXPECTED.$test_count
> +notmuch search --format=3Dtext0 --output=3Dthreads '*' | xargs -0 -L1 | =
notmuch_search_sanitize > OUTPUT.$test_count

I think it would be worth "strengthening" these tests as Mark pointed
out.  It would be easy to accidentally include a literal \n in the
output instead of calling the separator method, and this test wouldn't
catch that.  I think Mark's suggestion with tr is pretty good, since it
directly disambiguates \0 and \n in the output, while producing a
reasonable diff if things do go wrong.

> +test_expect_equal_file EXPECTED.$test_count OUTPUT.$test_count
> +
> +test_begin_subtest "Compare text vs. text0 for messages"
> +notmuch search --format=3Dtext --output=3Dmessages '*' | notmuch_search_=
sanitize > EXPECTED.$test_count
> +notmuch search --format=3Dtext0 --output=3Dmessages '*' | xargs -0 -L1 |=
 notmuch_search_sanitize > OUTPUT.$test_count
> +test_expect_equal_file EXPECTED.$test_count OUTPUT.$test_count
> +
> +test_begin_subtest "Compare text vs. text0 for files"
> +notmuch search --format=3Dtext --output=3Dfiles '*' | notmuch_search_san=
itize > EXPECTED.$test_count
> +notmuch search --format=3Dtext0 --output=3Dfiles '*' | xargs -0 -L1 | no=
tmuch_search_sanitize > OUTPUT.$test_count
> +test_expect_equal_file EXPECTED.$test_count OUTPUT.$test_count
> +
> +test_begin_subtest "Compare text vs. text0 for tags"
> +notmuch search --format=3Dtext --output=3Dtags '*' | notmuch_search_sani=
tize > EXPECTED.$test_count
> +notmuch search --format=3Dtext0 --output=3Dtags '*' | xargs -0 -L1 | not=
much_search_sanitize > OUTPUT.$test_count
> +test_expect_equal_file EXPECTED.$test_count OUTPUT.$test_count
> +
>  test_done
> --=20
> 1.7.10.4
>
> _______________________________________________
> notmuch mailing list
> notmuch@notmuchmail.org
> http://notmuchmail.org/mailman/listinfo/notmuch