Re: [PATCH 4/7] config: add --literal-value option, un-implemented

[Junio C Hamano <gitster@pobox.com>]

"Derrick Stolee via GitGitGadget" <gitgitgadget@gmail.com> writes:

> +--literal-value::
> +	When used with the `value_regex` argument, treat `value_regex` as
> +	an exact string instead of a regular expression. This will restrict
> +	the name/value pairs that are matched to only those where the value
> +	is exactly equal to the `value_regex`.

Makes sense.

>  --type <type>::
>    'git config' will ensure that any input or output is valid under the given
>    type constraint(s), and will canonicalize outgoing values in `<type>`'s
> diff --git a/builtin/config.c b/builtin/config.c
> index e7c7f3d455..ad6c695737 100644
> --- a/builtin/config.c
> +++ b/builtin/config.c
> @@ -34,6 +34,7 @@ static int respect_includes_opt = -1;
>  static struct config_options config_options;
>  static int show_origin;
>  static int show_scope;
> +static int literal;
>  
>  #define ACTION_GET (1<<0)
>  #define ACTION_GET_ALL (1<<1)
> @@ -52,6 +53,16 @@ static int show_scope;
>  #define ACTION_GET_COLORBOOL (1<<14)
>  #define ACTION_GET_URLMATCH (1<<15)
>  
> +#define ACTION_LITERAL_ALLOWED (\
> +	ACTION_GET |\
> +	ACTION_GET_ALL |\
> +	ACTION_GET_REGEXP |\
> +	ACTION_REPLACE_ALL |\
> +	ACTION_UNSET |\
> +	ACTION_UNSET_ALL |\
> +	ACTION_SET_ALL\
> +)

I am not sure about this, though.

Even if an action can potentially take value_regex, e.g.

    git config --get name [value_regex]

should we allow litral when value_regex is not given?  IOW

    $ git config --get --literal-value vari.able v2.26.0+next

may make sense, but shouldn't

    $ git config --get --literal-value vari.able

be an error?

To put it differently, I think the macro being defined is not
ACTION_LITERAL_ALLOWED, but ACTION_VALUE_REGEX_ALLOWED. i.e. list of
actions that can potentially take value_regex.  A command line that
gives value_regex to an action that is not listed there is an error.

> +	if (literal && !(actions & ACTION_LITERAL_ALLOWED)) {
> +		error(_("--literal only applies with 'value_regex'"));
> +		usage_builtin_config();
> +	}

This check is not incorrect per-se, because an action that never
takes value_regex is by definition incompatible with the literal
value option.  But it does not flag it as an error to ask a
value_regex to be treated as a literal string when there is no
value_regex given.

Having said that, I think it is OK, at least for now, to leave such
an error undetected---it falls into the "if it hurts, don't do it"
category.

Thanks.

>  	if (actions & PAGING_ACTIONS)
>  		setup_auto_pager("config", 1);
>  
> diff --git a/t/t1300-config.sh b/t/t1300-config.sh
> index 74e0f84c0a..73f5ca4361 100755
> --- a/t/t1300-config.sh
> +++ b/t/t1300-config.sh
> @@ -1965,4 +1965,17 @@ test_expect_success '--replace-all and value_regex' '
>  	test_cmp expect .git/config
>  '
>  
> +test_expect_success 'refuse --literal-value for incompatible actions' '
> +	git config dev.null bogus &&
> +	test_must_fail git config --literal-value --add dev.null bogus &&
> +	test_must_fail git config --literal-value --get-urlmatch dev.null bogus &&
> +	test_must_fail git config --literal-value --get-urlmatch dev.null bogus &&
> +	test_must_fail git config --literal-value --rename-section dev null &&
> +	test_must_fail git config --literal-value --remove-section dev &&
> +	test_must_fail git config --literal-value --list &&
> +	test_must_fail git config --literal-value --get-color dev.null &&
> +	test_must_fail git config --literal-value --get-colorbool dev.null &&
> +	test_must_fail git config --literal-value --edit
> +'
> +
>  test_done