Martin Ågren <martin.ag...@gmail.com> writes:

> +test_expect_success TTY 'git config respects pager.config when setting' '
> +     rm -f paginated.out &&
> +     test_terminal git -c pager.config config foo.bar bar &&
> +     test -e paginated.out
> +'

I am debating myself if this test should instead spell out what we
eventually want from the above test and make it expect_failure, just
like the next one.

In addition to setting (which will start ignoring pager in later
steps), unsetting, replacing of a variable and renaming/removing a
section in a configuration should not page, I would suspect.  Should
we test them all?

> +test_expect_failure TTY 'git config --edit ignores pager.config' '
> +     rm -f paginated.out editor.used &&
> +     write_script editor <<-\EOF &&
> +             touch editor.used
> +     EOF
> +     EDITOR=./editor test_terminal git -c pager.config config --edit &&
> +     ! test -e paginated.out &&
> +     test -e editor.used
> +'
> +
> +test_expect_success TTY 'git config --get defaults to not paging' '
> +     rm -f paginated.out &&
> +     test_terminal git config --get foo.bar &&
> +     ! test -e paginated.out
> +'
> +
> +test_expect_success TTY 'git config --get respects pager.config' '
> +     rm -f paginated.out &&
> +     test_terminal git -c pager.config config --get foo.bar &&
> +     test -e paginated.out
> +'
> +
> +test_expect_success TTY 'git config --list defaults to not paging' '
> +     rm -f paginated.out &&
> +     test_terminal git config --list &&
> +     ! test -e paginated.out
> +'
> +
> +
>  # A colored commit log will begin with an appropriate ANSI escape
>  # for the first color; the text "commit" comes later.
>  colorful() {

Reply via email to