[PATCH 3/3] test: implement and document NOTMUCH_TEST_QUIET variable usage
Austin Clements
amdragon at MIT.EDU
Tue Nov 12 15:02:59 PST 2013
Quoth Tomi Ollila on Nov 12 at 10:41 pm:
> When NOTMUCH_TEST_QUIET environment variable is set to non-null value
> messages when new test script starts and when test PASSes are disabled.
> This eases picking the cases when tests FAIL (as those are still printed).
> ---
> test/README | 8 ++++++++
> test/basic | 12 ++++++++++--
> test/test-lib.sh | 11 ++++++++++-
> test/test.expected-output/test-quiet-verbose-no | 20 ++++++++++++++++++++
> test/test.expected-output/test-quiet-verbose-yes | 24 ++++++++++++++++++++++++
> 5 files changed, 72 insertions(+), 3 deletions(-)
> create mode 100644 test/test.expected-output/test-quiet-verbose-no
> create mode 100644 test/test.expected-output/test-quiet-verbose-yes
>
> diff --git a/test/README b/test/README
> index d12cff2..79a9b1b 100644
> --- a/test/README
> +++ b/test/README
> @@ -76,6 +76,14 @@ the tests in one of the following ways.
> TEST_EMACS=my-special-emacs TEST_EMACSCLIENT=my-emacsclient ./emacs
> make test TEST_EMACS=my-special-emacs TEST_EMACSCLIENT=my-emacsclient
>
> +Quiet Execution
> +---------------
> +
> +Normally, when new script starts and when test PASSes you get a message
> +printed on screen. This printing can be disabled by setting the
> +NOTMUCH_TEST_QUIET variable to a non-null value. Message on test
> +failures and skips are still printed.
> +
> Skipping Tests
> --------------
> If, for any reason, you need to skip one or more tests, you can do so
> diff --git a/test/basic b/test/basic
> index 64eb7d7..3b7668b 100755
> --- a/test/basic
> +++ b/test/basic
> @@ -73,14 +73,22 @@ suppress_diff_date() {
> -e 's/\(.*\+\+\+ test-verbose\.4\.\output\).*/\1/'
> }
>
> +if [ -z "$NOTMUCH_TEST_QUIET" ]
> +then
> + test_verbose_no=$EXPECTED/test-verbose-no
> + test_verbose_yes=$EXPECTED/test-verbose-yes
> +else
> + test_verbose_no=$EXPECTED/test-quiet-verbose-no
> + test_verbose_yes=$EXPECTED/test-quiet-verbose-yes
> +fi
> test_begin_subtest "Ensure that test output is suppressed unless the test fails"
> output=$(cd $TEST_DIRECTORY; ./test-verbose 2>&1 | suppress_diff_date)
> -expected=$(cat $EXPECTED/test-verbose-no | suppress_diff_date)
> +expected=$(cat ${test_verbose_no} | suppress_diff_date)
> test_expect_equal "$output" "$expected"
>
> test_begin_subtest "Ensure that -v does not suppress test output"
> output=$(cd $TEST_DIRECTORY; ./test-verbose -v 2>&1 | suppress_diff_date)
> -expected=$(cat $EXPECTED/test-verbose-yes | suppress_diff_date)
> +expected=$(cat ${test_verbose_yes} | suppress_diff_date)
> # Do not include the results of test-verbose in totals
> rm $TEST_DIRECTORY/test-results/test-verbose
> rm -r $TEST_DIRECTORY/tmp.test-verbose
> diff --git a/test/test-lib.sh b/test/test-lib.sh
> index e022e46..4b342ac 100644
> --- a/test/test-lib.sh
> +++ b/test/test-lib.sh
> @@ -195,7 +195,10 @@ print_test_description ()
> echo
> echo $this_test: "Testing ${test_description}"
> }
> -print_test_description
> +if [ -z "$NOTMUCH_TEST_QUIET" ]
> +then
> + print_test_description
> +fi
>
> exec 5>&1
>
> @@ -703,6 +706,9 @@ test_ok_ () {
> return
> fi
> test_success=$(($test_success + 1))
> + if test -n "$NOTMUCH_TEST_QUIET"; then
> + return 0
> + fi
> say_color pass "%-6s" "PASS"
> echo " $test_subtest_name"
> }
> @@ -713,6 +719,9 @@ test_failure_ () {
> return
> fi
> test_failure=$(($test_failure + 1))
> + if test -n "$NOTMUCH_TEST_QUIET"; then
> + print_test_description
This prints the test description for *every* failing test. Was that
intentional? I would think that, ideally, it would be only printed
before the first failing subtest in a test (maybe by setting a
variable in print_test_description on the first call and making it
return immediately if this variable is set? Then you wouldn't even
need the condition here, just the call to print_test_description.)
> + fi
> test_failure_message_ "FAIL" "$test_subtest_name" "$@"
> test "$immediate" = "" || { GIT_EXIT_OK=t; exit 1; }
> return 1
Stylistic nit: The three if's above use two different styles ([ vs
test and hanging 'then'). OTOH, maybe this is consistent with
test-lib's inconsistent style.
> diff --git a/test/test.expected-output/test-quiet-verbose-no b/test/test.expected-output/test-quiet-verbose-no
> new file mode 100644
> index 0000000..74840b9
> --- /dev/null
> +++ b/test/test.expected-output/test-quiet-verbose-no
> @@ -0,0 +1,20 @@
> +
> +test-verbose: Testing the verbosity options of the test framework itself.
> + FAIL print something in test_expect_success and fail
> +
> + echo "hello stdout" &&
> + echo "hello stderr" >&2 &&
> + false
> +
> +hello stdout
> +hello stderr
> +
> +test-verbose: Testing the verbosity options of the test framework itself.
> + FAIL print something test_begin_subtest and test_expect_equal and fail
> + --- test-verbose.4.expected
> + +++ test-verbose.4.output
> + @@ -1 +1 @@
> + -b
> + +a
> +hello stdout
> +hello stderr
> diff --git a/test/test.expected-output/test-quiet-verbose-yes b/test/test.expected-output/test-quiet-verbose-yes
> new file mode 100644
> index 0000000..51e759d
> --- /dev/null
> +++ b/test/test.expected-output/test-quiet-verbose-yes
> @@ -0,0 +1,24 @@
> +hello stdout
> +hello stderr
> +hello stdout
> +hello stderr
> +
> +test-verbose: Testing the verbosity options of the test framework itself.
> + FAIL print something in test_expect_success and fail
> +
> + echo "hello stdout" &&
> + echo "hello stderr" >&2 &&
> + false
> +
> +hello stdout
> +hello stderr
> +hello stdout
> +hello stderr
> +
> +test-verbose: Testing the verbosity options of the test framework itself.
> + FAIL print something test_begin_subtest and test_expect_equal and fail
> + --- test-verbose.4.expected
> + +++ test-verbose.4.output
> + @@ -1 +1 @@
> + -b
> + +a
More information about the notmuch
mailing list