2007-10-15 07:35:37 +08:00
|
|
|
#ifndef PARSE_OPTIONS_H
|
|
|
|
#define PARSE_OPTIONS_H
|
|
|
|
|
|
|
|
enum parse_opt_type {
|
2007-11-07 18:20:27 +08:00
|
|
|
/* special types */
|
2007-10-15 07:35:37 +08:00
|
|
|
OPTION_END,
|
2008-03-02 18:35:56 +08:00
|
|
|
OPTION_ARGUMENT,
|
2007-10-15 07:38:30 +08:00
|
|
|
OPTION_GROUP,
|
2009-05-08 03:45:08 +08:00
|
|
|
OPTION_NUMBER,
|
2007-11-07 18:20:27 +08:00
|
|
|
/* options with no arguments */
|
|
|
|
OPTION_BIT,
|
2009-05-08 03:44:17 +08:00
|
|
|
OPTION_NEGBIT,
|
parse-options: deprecate OPT_BOOLEAN
It is natural to expect that an option defined with OPT_BOOLEAN() could be
used in this way:
int option = -1; /* unspecified */
struct option options[] = {
OPT_BOOLEAN(0, "option", &option, "set option"),
OPT_END()
};
parse_options(ac, av, prefix, options, usage, 0);
if (option < 0)
... do the default thing ...
else if (!option)
... --no-option was given ...
else
... --option was given ...
to easily tell three cases apart:
- There is no mention of the `--option` on the command line;
- The variable is positively set with `--option`; or
- The variable is explicitly negated with `--no-option`.
Unfortunately, this is not the case. OPT_BOOLEAN() increments the variable
every time `--option` is given, and resets it to zero when `--no-option`
is given.
As a first step to remedy this, introduce a true boolean OPT_BOOL(), and
rename OPT_BOOLEAN() to OPT_COUNTUP(). To help transitioning, OPT_BOOLEAN
and OPTION_BOOLEAN are defined as deprecated synonyms to OPT_COUNTUP and
OPTION_COUNTUP respectively.
This is what db7244b (parse-options new features., 2007-11-07) from four
years ago started by marking OPTION_BOOLEAN as "INCR would have been a
better name".
Some existing users do depend on the count-up semantics; for example,
users of OPT__VERBOSE() could use it to raise the verbosity level with
repeated use of `-v` on the command line, but they probably should be
rewritten to use OPT__VERBOSITY() instead these days. I suspect that some
users of OPT__FORCE() may also use it to implement different level of
forcibleness but I didn't check.
On top of this patch, here are the remaining clean-up tasks that other
people can help:
- Look at each hit in "git grep -e OPT_BOOLEAN"; trace all uses of the
value that is set to the underlying variable, and if it can proven that
the variable is only used as a boolean, replace it with OPT_BOOL(). If
the caller does depend on the count-up semantics, replace it with
OPT_COUNTUP() instead.
- Same for OPTION_BOOLEAN; replace it with OPTION_SET_INT and arrange to
set 1 to the variable for a true boolean, and otherwise replace it with
OPTION_COUNTUP.
- Look at each hit in "git grep -e OPT__VERBOSE -e OPT__QUIET" and see if
they can be replaced with OPT__VERBOSITY().
I'll follow this message up with a separate patch as an example.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2011-09-28 07:56:49 +08:00
|
|
|
OPTION_COUNTUP,
|
2007-11-07 18:20:27 +08:00
|
|
|
OPTION_SET_INT,
|
|
|
|
OPTION_SET_PTR,
|
|
|
|
/* options with arguments (usually) */
|
2007-10-15 07:35:37 +08:00
|
|
|
OPTION_STRING,
|
|
|
|
OPTION_INTEGER,
|
2007-10-15 07:45:45 +08:00
|
|
|
OPTION_CALLBACK,
|
parse-options: allow git commands to invent new option types
parse-options provides a variety of option behaviors, including
OPTION_CALLBACK, which should take care of just about any sane
behavior. All supported behaviors obey the following constraint:
A --foo option can only accept (and base its behavior on)
one argument, which would be the following command-line
argument in the "unsticked" form.
Alas, some existing git commands have options that do not obey that
constraint. For example, update-index --cacheinfo takes three
arguments, and update-index --resolve takes all later parameters as
arguments.
Introduces an OPTION_LOWLEVEL_CALLBACK backdoor to parse-options so
such option types can be supported without tempting inventors of other
commands through mention in the public API. Commands can set the
callback field to a function accepting three arguments: the option
parsing context, the option itself, and a flag indicating whether the
the option was negated. When the option is encountered, that function
is called to take over from get_value(). The return value should be
zero for success, -1 for usage errors.
Thanks to Stephen Boyd for API guidance.
Improved-by: Stephen Boyd <bebarino@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2010-12-02 07:32:16 +08:00
|
|
|
OPTION_LOWLEVEL_CALLBACK,
|
2009-05-24 02:53:13 +08:00
|
|
|
OPTION_FILENAME
|
2007-10-15 07:35:37 +08:00
|
|
|
};
|
|
|
|
|
parse-options: deprecate OPT_BOOLEAN
It is natural to expect that an option defined with OPT_BOOLEAN() could be
used in this way:
int option = -1; /* unspecified */
struct option options[] = {
OPT_BOOLEAN(0, "option", &option, "set option"),
OPT_END()
};
parse_options(ac, av, prefix, options, usage, 0);
if (option < 0)
... do the default thing ...
else if (!option)
... --no-option was given ...
else
... --option was given ...
to easily tell three cases apart:
- There is no mention of the `--option` on the command line;
- The variable is positively set with `--option`; or
- The variable is explicitly negated with `--no-option`.
Unfortunately, this is not the case. OPT_BOOLEAN() increments the variable
every time `--option` is given, and resets it to zero when `--no-option`
is given.
As a first step to remedy this, introduce a true boolean OPT_BOOL(), and
rename OPT_BOOLEAN() to OPT_COUNTUP(). To help transitioning, OPT_BOOLEAN
and OPTION_BOOLEAN are defined as deprecated synonyms to OPT_COUNTUP and
OPTION_COUNTUP respectively.
This is what db7244b (parse-options new features., 2007-11-07) from four
years ago started by marking OPTION_BOOLEAN as "INCR would have been a
better name".
Some existing users do depend on the count-up semantics; for example,
users of OPT__VERBOSE() could use it to raise the verbosity level with
repeated use of `-v` on the command line, but they probably should be
rewritten to use OPT__VERBOSITY() instead these days. I suspect that some
users of OPT__FORCE() may also use it to implement different level of
forcibleness but I didn't check.
On top of this patch, here are the remaining clean-up tasks that other
people can help:
- Look at each hit in "git grep -e OPT_BOOLEAN"; trace all uses of the
value that is set to the underlying variable, and if it can proven that
the variable is only used as a boolean, replace it with OPT_BOOL(). If
the caller does depend on the count-up semantics, replace it with
OPT_COUNTUP() instead.
- Same for OPTION_BOOLEAN; replace it with OPTION_SET_INT and arrange to
set 1 to the variable for a true boolean, and otherwise replace it with
OPTION_COUNTUP.
- Look at each hit in "git grep -e OPT__VERBOSE -e OPT__QUIET" and see if
they can be replaced with OPT__VERBOSITY().
I'll follow this message up with a separate patch as an example.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2011-09-28 07:56:49 +08:00
|
|
|
/* Deprecated synonym */
|
|
|
|
#define OPTION_BOOLEAN OPTION_COUNTUP
|
|
|
|
|
2007-10-15 07:35:37 +08:00
|
|
|
enum parse_opt_flags {
|
|
|
|
PARSE_OPT_KEEP_DASHDASH = 1,
|
2008-02-29 09:45:09 +08:00
|
|
|
PARSE_OPT_STOP_AT_NON_OPTION = 2,
|
2008-06-24 06:31:31 +08:00
|
|
|
PARSE_OPT_KEEP_ARGV0 = 4,
|
2009-03-09 02:12:47 +08:00
|
|
|
PARSE_OPT_KEEP_UNKNOWN = 8,
|
2010-05-14 17:31:35 +08:00
|
|
|
PARSE_OPT_NO_INTERNAL_HELP = 16
|
2007-10-15 07:35:37 +08:00
|
|
|
};
|
|
|
|
|
2007-10-15 07:45:45 +08:00
|
|
|
enum parse_opt_option_flags {
|
|
|
|
PARSE_OPT_OPTARG = 1,
|
2007-10-16 06:32:38 +08:00
|
|
|
PARSE_OPT_NOARG = 2,
|
2007-11-07 18:20:27 +08:00
|
|
|
PARSE_OPT_NONEG = 4,
|
2007-11-19 17:21:44 +08:00
|
|
|
PARSE_OPT_HIDDEN = 8,
|
2008-07-08 18:34:08 +08:00
|
|
|
PARSE_OPT_LASTARG_DEFAULT = 16,
|
2009-05-08 03:45:42 +08:00
|
|
|
PARSE_OPT_NODASH = 32,
|
2009-05-21 15:33:17 +08:00
|
|
|
PARSE_OPT_LITERAL_ARGHELP = 64,
|
2009-07-28 02:49:56 +08:00
|
|
|
PARSE_OPT_NEGHELP = 128,
|
2010-06-12 20:57:39 +08:00
|
|
|
PARSE_OPT_SHELL_EVAL = 256
|
2007-10-15 07:45:45 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
struct option;
|
|
|
|
typedef int parse_opt_cb(const struct option *, const char *arg, int unset);
|
|
|
|
|
parse-options: allow git commands to invent new option types
parse-options provides a variety of option behaviors, including
OPTION_CALLBACK, which should take care of just about any sane
behavior. All supported behaviors obey the following constraint:
A --foo option can only accept (and base its behavior on)
one argument, which would be the following command-line
argument in the "unsticked" form.
Alas, some existing git commands have options that do not obey that
constraint. For example, update-index --cacheinfo takes three
arguments, and update-index --resolve takes all later parameters as
arguments.
Introduces an OPTION_LOWLEVEL_CALLBACK backdoor to parse-options so
such option types can be supported without tempting inventors of other
commands through mention in the public API. Commands can set the
callback field to a function accepting three arguments: the option
parsing context, the option itself, and a flag indicating whether the
the option was negated. When the option is encountered, that function
is called to take over from get_value(). The return value should be
zero for success, -1 for usage errors.
Thanks to Stephen Boyd for API guidance.
Improved-by: Stephen Boyd <bebarino@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2010-12-02 07:32:16 +08:00
|
|
|
struct parse_opt_ctx_t;
|
|
|
|
typedef int parse_opt_ll_cb(struct parse_opt_ctx_t *ctx,
|
|
|
|
const struct option *opt, int unset);
|
|
|
|
|
2007-11-05 20:03:22 +08:00
|
|
|
/*
|
|
|
|
* `type`::
|
|
|
|
* holds the type of the option, you must have an OPTION_END last in your
|
|
|
|
* array.
|
|
|
|
*
|
|
|
|
* `short_name`::
|
|
|
|
* the character to use as a short option name, '\0' if none.
|
|
|
|
*
|
|
|
|
* `long_name`::
|
|
|
|
* the long option name, without the leading dashes, NULL if none.
|
|
|
|
*
|
|
|
|
* `value`::
|
|
|
|
* stores pointers to the values to be filled.
|
|
|
|
*
|
|
|
|
* `argh`::
|
|
|
|
* token to explain the kind of argument this option wants. Keep it
|
2009-04-18 02:13:30 +08:00
|
|
|
* homogeneous across the repository.
|
2007-11-05 20:03:22 +08:00
|
|
|
*
|
|
|
|
* `help`::
|
|
|
|
* the short help associated to what the option does.
|
|
|
|
* Must never be NULL (except for OPTION_END).
|
|
|
|
* OPTION_GROUP uses this pointer to store the group header.
|
|
|
|
*
|
|
|
|
* `flags`::
|
|
|
|
* mask of parse_opt_option_flags.
|
2009-04-18 02:13:30 +08:00
|
|
|
* PARSE_OPT_OPTARG: says that the argument is optional (not for BOOLEANs)
|
2010-08-23 00:26:38 +08:00
|
|
|
* PARSE_OPT_NOARG: says that this option does not take an argument
|
2007-11-07 18:20:27 +08:00
|
|
|
* PARSE_OPT_NONEG: says that this option cannot be negated
|
2009-05-08 03:45:42 +08:00
|
|
|
* PARSE_OPT_HIDDEN: this option is skipped in the default usage, and
|
|
|
|
* shown only in the full usage.
|
2009-06-08 07:39:15 +08:00
|
|
|
* PARSE_OPT_LASTARG_DEFAULT: says that this option will take the default
|
|
|
|
* value if no argument is given when the option
|
|
|
|
* is last on the command line. If the option is
|
|
|
|
* not last it will require an argument.
|
|
|
|
* Should not be used with PARSE_OPT_OPTARG.
|
2009-05-08 03:45:42 +08:00
|
|
|
* PARSE_OPT_NODASH: this option doesn't start with a dash.
|
2009-05-21 15:33:17 +08:00
|
|
|
* PARSE_OPT_LITERAL_ARGHELP: says that argh shouldn't be enclosed in brackets
|
|
|
|
* (i.e. '<argh>') in the help message.
|
|
|
|
* Useful for options with multiple parameters.
|
2009-07-28 02:49:56 +08:00
|
|
|
* PARSE_OPT_NEGHELP: says that the long option should always be shown with
|
|
|
|
* the --no prefix in the usage message. Sometimes
|
|
|
|
* useful for users of OPTION_NEGBIT.
|
2007-11-05 20:03:22 +08:00
|
|
|
*
|
|
|
|
* `callback`::
|
parse-options: allow git commands to invent new option types
parse-options provides a variety of option behaviors, including
OPTION_CALLBACK, which should take care of just about any sane
behavior. All supported behaviors obey the following constraint:
A --foo option can only accept (and base its behavior on)
one argument, which would be the following command-line
argument in the "unsticked" form.
Alas, some existing git commands have options that do not obey that
constraint. For example, update-index --cacheinfo takes three
arguments, and update-index --resolve takes all later parameters as
arguments.
Introduces an OPTION_LOWLEVEL_CALLBACK backdoor to parse-options so
such option types can be supported without tempting inventors of other
commands through mention in the public API. Commands can set the
callback field to a function accepting three arguments: the option
parsing context, the option itself, and a flag indicating whether the
the option was negated. When the option is encountered, that function
is called to take over from get_value(). The return value should be
zero for success, -1 for usage errors.
Thanks to Stephen Boyd for API guidance.
Improved-by: Stephen Boyd <bebarino@gmail.com>
Signed-off-by: Jonathan Nieder <jrnieder@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2010-12-02 07:32:16 +08:00
|
|
|
* pointer to the callback to use for OPTION_CALLBACK or
|
|
|
|
* OPTION_LOWLEVEL_CALLBACK.
|
2007-11-05 20:03:22 +08:00
|
|
|
*
|
|
|
|
* `defval`::
|
|
|
|
* default value to fill (*->value) with for PARSE_OPT_OPTARG.
|
2007-11-07 18:20:27 +08:00
|
|
|
* OPTION_{BIT,SET_INT,SET_PTR} store the {mask,integer,pointer} to put in
|
|
|
|
* the value when met.
|
2007-11-05 20:03:22 +08:00
|
|
|
* CALLBACKS can use it like they want.
|
|
|
|
*/
|
2007-10-15 07:35:37 +08:00
|
|
|
struct option {
|
|
|
|
enum parse_opt_type type;
|
|
|
|
int short_name;
|
|
|
|
const char *long_name;
|
|
|
|
void *value;
|
2007-10-15 07:38:30 +08:00
|
|
|
const char *argh;
|
|
|
|
const char *help;
|
2007-10-15 07:45:45 +08:00
|
|
|
|
|
|
|
int flags;
|
|
|
|
parse_opt_cb *callback;
|
|
|
|
intptr_t defval;
|
2007-10-15 07:35:37 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
#define OPT_END() { OPTION_END }
|
2009-06-05 07:43:57 +08:00
|
|
|
#define OPT_ARGUMENT(l, h) { OPTION_ARGUMENT, 0, (l), NULL, NULL, \
|
|
|
|
(h), PARSE_OPT_NOARG}
|
2007-10-15 07:38:30 +08:00
|
|
|
#define OPT_GROUP(h) { OPTION_GROUP, 0, NULL, NULL, NULL, (h) }
|
2009-06-05 07:43:57 +08:00
|
|
|
#define OPT_BIT(s, l, v, h, b) { OPTION_BIT, (s), (l), (v), NULL, (h), \
|
|
|
|
PARSE_OPT_NOARG, NULL, (b) }
|
|
|
|
#define OPT_NEGBIT(s, l, v, h, b) { OPTION_NEGBIT, (s), (l), (v), NULL, \
|
|
|
|
(h), PARSE_OPT_NOARG, NULL, (b) }
|
parse-options: deprecate OPT_BOOLEAN
It is natural to expect that an option defined with OPT_BOOLEAN() could be
used in this way:
int option = -1; /* unspecified */
struct option options[] = {
OPT_BOOLEAN(0, "option", &option, "set option"),
OPT_END()
};
parse_options(ac, av, prefix, options, usage, 0);
if (option < 0)
... do the default thing ...
else if (!option)
... --no-option was given ...
else
... --option was given ...
to easily tell three cases apart:
- There is no mention of the `--option` on the command line;
- The variable is positively set with `--option`; or
- The variable is explicitly negated with `--no-option`.
Unfortunately, this is not the case. OPT_BOOLEAN() increments the variable
every time `--option` is given, and resets it to zero when `--no-option`
is given.
As a first step to remedy this, introduce a true boolean OPT_BOOL(), and
rename OPT_BOOLEAN() to OPT_COUNTUP(). To help transitioning, OPT_BOOLEAN
and OPTION_BOOLEAN are defined as deprecated synonyms to OPT_COUNTUP and
OPTION_COUNTUP respectively.
This is what db7244b (parse-options new features., 2007-11-07) from four
years ago started by marking OPTION_BOOLEAN as "INCR would have been a
better name".
Some existing users do depend on the count-up semantics; for example,
users of OPT__VERBOSE() could use it to raise the verbosity level with
repeated use of `-v` on the command line, but they probably should be
rewritten to use OPT__VERBOSITY() instead these days. I suspect that some
users of OPT__FORCE() may also use it to implement different level of
forcibleness but I didn't check.
On top of this patch, here are the remaining clean-up tasks that other
people can help:
- Look at each hit in "git grep -e OPT_BOOLEAN"; trace all uses of the
value that is set to the underlying variable, and if it can proven that
the variable is only used as a boolean, replace it with OPT_BOOL(). If
the caller does depend on the count-up semantics, replace it with
OPT_COUNTUP() instead.
- Same for OPTION_BOOLEAN; replace it with OPTION_SET_INT and arrange to
set 1 to the variable for a true boolean, and otherwise replace it with
OPTION_COUNTUP.
- Look at each hit in "git grep -e OPT__VERBOSE -e OPT__QUIET" and see if
they can be replaced with OPT__VERBOSITY().
I'll follow this message up with a separate patch as an example.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2011-09-28 07:56:49 +08:00
|
|
|
#define OPT_COUNTUP(s, l, v, h) { OPTION_COUNTUP, (s), (l), (v), NULL, \
|
2009-06-05 07:43:57 +08:00
|
|
|
(h), PARSE_OPT_NOARG }
|
|
|
|
#define OPT_SET_INT(s, l, v, h, i) { OPTION_SET_INT, (s), (l), (v), NULL, \
|
|
|
|
(h), PARSE_OPT_NOARG, NULL, (i) }
|
parse-options: deprecate OPT_BOOLEAN
It is natural to expect that an option defined with OPT_BOOLEAN() could be
used in this way:
int option = -1; /* unspecified */
struct option options[] = {
OPT_BOOLEAN(0, "option", &option, "set option"),
OPT_END()
};
parse_options(ac, av, prefix, options, usage, 0);
if (option < 0)
... do the default thing ...
else if (!option)
... --no-option was given ...
else
... --option was given ...
to easily tell three cases apart:
- There is no mention of the `--option` on the command line;
- The variable is positively set with `--option`; or
- The variable is explicitly negated with `--no-option`.
Unfortunately, this is not the case. OPT_BOOLEAN() increments the variable
every time `--option` is given, and resets it to zero when `--no-option`
is given.
As a first step to remedy this, introduce a true boolean OPT_BOOL(), and
rename OPT_BOOLEAN() to OPT_COUNTUP(). To help transitioning, OPT_BOOLEAN
and OPTION_BOOLEAN are defined as deprecated synonyms to OPT_COUNTUP and
OPTION_COUNTUP respectively.
This is what db7244b (parse-options new features., 2007-11-07) from four
years ago started by marking OPTION_BOOLEAN as "INCR would have been a
better name".
Some existing users do depend on the count-up semantics; for example,
users of OPT__VERBOSE() could use it to raise the verbosity level with
repeated use of `-v` on the command line, but they probably should be
rewritten to use OPT__VERBOSITY() instead these days. I suspect that some
users of OPT__FORCE() may also use it to implement different level of
forcibleness but I didn't check.
On top of this patch, here are the remaining clean-up tasks that other
people can help:
- Look at each hit in "git grep -e OPT_BOOLEAN"; trace all uses of the
value that is set to the underlying variable, and if it can proven that
the variable is only used as a boolean, replace it with OPT_BOOL(). If
the caller does depend on the count-up semantics, replace it with
OPT_COUNTUP() instead.
- Same for OPTION_BOOLEAN; replace it with OPTION_SET_INT and arrange to
set 1 to the variable for a true boolean, and otherwise replace it with
OPTION_COUNTUP.
- Look at each hit in "git grep -e OPT__VERBOSE -e OPT__QUIET" and see if
they can be replaced with OPT__VERBOSITY().
I'll follow this message up with a separate patch as an example.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2011-09-28 07:56:49 +08:00
|
|
|
#define OPT_BOOL(s, l, v, h) OPT_SET_INT(s, l, v, h, 1)
|
2009-06-05 07:43:57 +08:00
|
|
|
#define OPT_SET_PTR(s, l, v, h, p) { OPTION_SET_PTR, (s), (l), (v), NULL, \
|
|
|
|
(h), PARSE_OPT_NOARG, NULL, (p) }
|
2009-06-05 07:43:56 +08:00
|
|
|
#define OPT_INTEGER(s, l, v, h) { OPTION_INTEGER, (s), (l), (v), "n", (h) }
|
2007-10-15 07:38:30 +08:00
|
|
|
#define OPT_STRING(s, l, v, a, h) { OPTION_STRING, (s), (l), (v), (a), (h) }
|
2011-06-09 23:55:23 +08:00
|
|
|
#define OPT_STRING_LIST(s, l, v, a, h) \
|
|
|
|
{ OPTION_CALLBACK, (s), (l), (v), (a), \
|
|
|
|
(h), 0, &parse_opt_string_list }
|
2009-12-04 16:20:48 +08:00
|
|
|
#define OPT_UYN(s, l, v, h) { OPTION_CALLBACK, (s), (l), (v), NULL, \
|
|
|
|
(h), PARSE_OPT_NOARG, &parse_opt_tertiary }
|
2008-03-24 22:02:21 +08:00
|
|
|
#define OPT_DATE(s, l, v, h) \
|
|
|
|
{ OPTION_CALLBACK, (s), (l), (v), "time",(h), 0, \
|
|
|
|
parse_opt_approxidate_cb }
|
2007-10-15 07:45:45 +08:00
|
|
|
#define OPT_CALLBACK(s, l, v, a, h, f) \
|
|
|
|
{ OPTION_CALLBACK, (s), (l), (v), (a), (h), 0, (f) }
|
2009-05-08 03:45:08 +08:00
|
|
|
#define OPT_NUMBER_CALLBACK(v, h, f) \
|
|
|
|
{ OPTION_NUMBER, 0, NULL, (v), NULL, (h), \
|
|
|
|
PARSE_OPT_NOARG | PARSE_OPT_NONEG, (f) }
|
2009-05-24 02:53:13 +08:00
|
|
|
#define OPT_FILENAME(s, l, v, h) { OPTION_FILENAME, (s), (l), (v), \
|
2011-02-15 21:09:11 +08:00
|
|
|
"file", (h) }
|
Add an optional argument for --color options
Make git-branch, git-show-branch, git-grep, and all the diff-based
programs accept an optional argument <when> for --color. The argument
is a colorbool: "always", "never", or "auto". If no argument is given,
"always" is used; --no-color is an alias for --color=never. This makes
the command-line interface consistent with other GNU tools, such as `ls'
and `grep', and with the git-config color options. Note that, without
an argument, --color and --no-color work exactly as before.
To implement this, two internal changes were made:
1. Allow the first argument of git_config_colorbool() to be NULL,
in which case it returns -1 if the argument isn't "always", "never",
or "auto".
2. Add OPT_COLOR_FLAG(), OPT__COLOR(), and parse_opt_color_flag_cb()
to the option parsing library. The callback uses
git_config_colorbool(), so color.h is now a dependency
of parse-options.c.
Signed-off-by: Mark Lodato <lodatom@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2010-02-17 12:55:58 +08:00
|
|
|
#define OPT_COLOR_FLAG(s, l, v, h) \
|
|
|
|
{ OPTION_CALLBACK, (s), (l), (v), "when", (h), PARSE_OPT_OPTARG, \
|
|
|
|
parse_opt_color_flag_cb, (intptr_t)"always" }
|
|
|
|
|
2011-09-29 01:44:30 +08:00
|
|
|
#define OPT_NOOP_NOARG(s, l) \
|
|
|
|
{ OPTION_CALLBACK, (s), (l), NULL, NULL, \
|
|
|
|
"no-op (backward compatibility)", \
|
|
|
|
PARSE_OPT_HIDDEN | PARSE_OPT_NOARG, parse_opt_noop_cb }
|
|
|
|
|
parse-options: deprecate OPT_BOOLEAN
It is natural to expect that an option defined with OPT_BOOLEAN() could be
used in this way:
int option = -1; /* unspecified */
struct option options[] = {
OPT_BOOLEAN(0, "option", &option, "set option"),
OPT_END()
};
parse_options(ac, av, prefix, options, usage, 0);
if (option < 0)
... do the default thing ...
else if (!option)
... --no-option was given ...
else
... --option was given ...
to easily tell three cases apart:
- There is no mention of the `--option` on the command line;
- The variable is positively set with `--option`; or
- The variable is explicitly negated with `--no-option`.
Unfortunately, this is not the case. OPT_BOOLEAN() increments the variable
every time `--option` is given, and resets it to zero when `--no-option`
is given.
As a first step to remedy this, introduce a true boolean OPT_BOOL(), and
rename OPT_BOOLEAN() to OPT_COUNTUP(). To help transitioning, OPT_BOOLEAN
and OPTION_BOOLEAN are defined as deprecated synonyms to OPT_COUNTUP and
OPTION_COUNTUP respectively.
This is what db7244b (parse-options new features., 2007-11-07) from four
years ago started by marking OPTION_BOOLEAN as "INCR would have been a
better name".
Some existing users do depend on the count-up semantics; for example,
users of OPT__VERBOSE() could use it to raise the verbosity level with
repeated use of `-v` on the command line, but they probably should be
rewritten to use OPT__VERBOSITY() instead these days. I suspect that some
users of OPT__FORCE() may also use it to implement different level of
forcibleness but I didn't check.
On top of this patch, here are the remaining clean-up tasks that other
people can help:
- Look at each hit in "git grep -e OPT_BOOLEAN"; trace all uses of the
value that is set to the underlying variable, and if it can proven that
the variable is only used as a boolean, replace it with OPT_BOOL(). If
the caller does depend on the count-up semantics, replace it with
OPT_COUNTUP() instead.
- Same for OPTION_BOOLEAN; replace it with OPTION_SET_INT and arrange to
set 1 to the variable for a true boolean, and otherwise replace it with
OPTION_COUNTUP.
- Look at each hit in "git grep -e OPT__VERBOSE -e OPT__QUIET" and see if
they can be replaced with OPT__VERBOSITY().
I'll follow this message up with a separate patch as an example.
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2011-09-28 07:56:49 +08:00
|
|
|
/* Deprecated synonym */
|
|
|
|
#define OPT_BOOLEAN OPT_COUNTUP
|
2007-10-15 07:35:37 +08:00
|
|
|
|
|
|
|
/* parse_options() will filter out the processed options and leave the
|
2009-04-18 02:13:30 +08:00
|
|
|
* non-option arguments in argv[].
|
2007-10-15 07:35:37 +08:00
|
|
|
* Returns the number of arguments left in argv[].
|
|
|
|
*/
|
2009-05-24 02:53:12 +08:00
|
|
|
extern int parse_options(int argc, const char **argv, const char *prefix,
|
2007-10-15 07:35:37 +08:00
|
|
|
const struct option *options,
|
2007-10-15 07:38:30 +08:00
|
|
|
const char * const usagestr[], int flags);
|
|
|
|
|
|
|
|
extern NORETURN void usage_with_options(const char * const *usagestr,
|
|
|
|
const struct option *options);
|
2007-10-15 07:35:37 +08:00
|
|
|
|
2009-02-02 13:12:58 +08:00
|
|
|
extern NORETURN void usage_msg_opt(const char *msg,
|
|
|
|
const char * const *usagestr,
|
|
|
|
const struct option *options);
|
|
|
|
|
2011-08-11 17:15:37 +08:00
|
|
|
extern int optbug(const struct option *opt, const char *reason);
|
|
|
|
extern int opterror(const struct option *opt, const char *reason, int flags);
|
2009-04-18 02:13:30 +08:00
|
|
|
/*----- incremental advanced APIs -----*/
|
2008-06-24 03:59:37 +08:00
|
|
|
|
2008-06-24 04:28:04 +08:00
|
|
|
enum {
|
|
|
|
PARSE_OPT_HELP = -1,
|
|
|
|
PARSE_OPT_DONE,
|
2010-12-02 07:32:55 +08:00
|
|
|
PARSE_OPT_NON_OPTION,
|
2010-05-14 17:31:35 +08:00
|
|
|
PARSE_OPT_UNKNOWN
|
2008-06-24 04:28:04 +08:00
|
|
|
};
|
|
|
|
|
2008-06-24 04:55:11 +08:00
|
|
|
/*
|
|
|
|
* It's okay for the caller to consume argv/argc in the usual way.
|
|
|
|
* Other fields of that structure are private to parse-options and should not
|
|
|
|
* be modified in any way.
|
|
|
|
*/
|
2008-06-24 03:59:37 +08:00
|
|
|
struct parse_opt_ctx_t {
|
|
|
|
const char **argv;
|
|
|
|
const char **out;
|
|
|
|
int argc, cpidx;
|
|
|
|
const char *opt;
|
|
|
|
int flags;
|
2009-05-24 02:53:12 +08:00
|
|
|
const char *prefix;
|
2008-06-24 03:59:37 +08:00
|
|
|
};
|
|
|
|
|
|
|
|
extern void parse_options_start(struct parse_opt_ctx_t *ctx,
|
2009-05-24 02:53:12 +08:00
|
|
|
int argc, const char **argv, const char *prefix,
|
2010-12-06 15:57:42 +08:00
|
|
|
const struct option *options, int flags);
|
2008-06-24 03:59:37 +08:00
|
|
|
|
2008-06-24 04:38:58 +08:00
|
|
|
extern int parse_options_step(struct parse_opt_ctx_t *ctx,
|
|
|
|
const struct option *options,
|
|
|
|
const char * const usagestr[]);
|
|
|
|
|
2008-06-24 03:59:37 +08:00
|
|
|
extern int parse_options_end(struct parse_opt_ctx_t *ctx);
|
|
|
|
|
2010-03-07 04:34:40 +08:00
|
|
|
extern int parse_options_concat(struct option *dst, size_t, struct option *src);
|
2008-06-24 03:59:37 +08:00
|
|
|
|
2007-10-14 17:05:12 +08:00
|
|
|
/*----- some often used options -----*/
|
|
|
|
extern int parse_opt_abbrev_cb(const struct option *, const char *, int);
|
2008-03-24 22:02:21 +08:00
|
|
|
extern int parse_opt_approxidate_cb(const struct option *, const char *, int);
|
Add an optional argument for --color options
Make git-branch, git-show-branch, git-grep, and all the diff-based
programs accept an optional argument <when> for --color. The argument
is a colorbool: "always", "never", or "auto". If no argument is given,
"always" is used; --no-color is an alias for --color=never. This makes
the command-line interface consistent with other GNU tools, such as `ls'
and `grep', and with the git-config color options. Note that, without
an argument, --color and --no-color work exactly as before.
To implement this, two internal changes were made:
1. Allow the first argument of git_config_colorbool() to be NULL,
in which case it returns -1 if the argument isn't "always", "never",
or "auto".
2. Add OPT_COLOR_FLAG(), OPT__COLOR(), and parse_opt_color_flag_cb()
to the option parsing library. The callback uses
git_config_colorbool(), so color.h is now a dependency
of parse-options.c.
Signed-off-by: Mark Lodato <lodatom@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2010-02-17 12:55:58 +08:00
|
|
|
extern int parse_opt_color_flag_cb(const struct option *, const char *, int);
|
2008-11-15 08:14:24 +08:00
|
|
|
extern int parse_opt_verbosity_cb(const struct option *, const char *, int);
|
2009-01-26 22:13:23 +08:00
|
|
|
extern int parse_opt_with_commit(const struct option *, const char *, int);
|
2009-12-04 16:20:48 +08:00
|
|
|
extern int parse_opt_tertiary(const struct option *, const char *, int);
|
2011-06-09 23:55:23 +08:00
|
|
|
extern int parse_opt_string_list(const struct option *, const char *, int);
|
2011-09-29 01:44:30 +08:00
|
|
|
extern int parse_opt_noop_cb(const struct option *, const char *, int);
|
2007-10-14 17:05:12 +08:00
|
|
|
|
2010-11-09 01:56:39 +08:00
|
|
|
#define OPT__VERBOSE(var, h) OPT_BOOLEAN('v', "verbose", (var), (h))
|
2010-11-09 02:06:54 +08:00
|
|
|
#define OPT__QUIET(var, h) OPT_BOOLEAN('q', "quiet", (var), (h))
|
2008-11-15 08:14:24 +08:00
|
|
|
#define OPT__VERBOSITY(var) \
|
|
|
|
{ OPTION_CALLBACK, 'v', "verbose", (var), NULL, "be more verbose", \
|
|
|
|
PARSE_OPT_NOARG, &parse_opt_verbosity_cb, 0 }, \
|
|
|
|
{ OPTION_CALLBACK, 'q', "quiet", (var), NULL, "be more quiet", \
|
|
|
|
PARSE_OPT_NOARG, &parse_opt_verbosity_cb, 0 }
|
2010-11-09 01:58:51 +08:00
|
|
|
#define OPT__DRY_RUN(var, h) OPT_BOOLEAN('n', "dry-run", (var), (h))
|
2010-11-09 02:01:54 +08:00
|
|
|
#define OPT__FORCE(var, h) OPT_BOOLEAN('f', "force", (var), (h))
|
2007-10-14 17:05:12 +08:00
|
|
|
#define OPT__ABBREV(var) \
|
|
|
|
{ OPTION_CALLBACK, 0, "abbrev", (var), "n", \
|
|
|
|
"use <n> digits to display SHA-1s", \
|
|
|
|
PARSE_OPT_OPTARG, &parse_opt_abbrev_cb, 0 }
|
Add an optional argument for --color options
Make git-branch, git-show-branch, git-grep, and all the diff-based
programs accept an optional argument <when> for --color. The argument
is a colorbool: "always", "never", or "auto". If no argument is given,
"always" is used; --no-color is an alias for --color=never. This makes
the command-line interface consistent with other GNU tools, such as `ls'
and `grep', and with the git-config color options. Note that, without
an argument, --color and --no-color work exactly as before.
To implement this, two internal changes were made:
1. Allow the first argument of git_config_colorbool() to be NULL,
in which case it returns -1 if the argument isn't "always", "never",
or "auto".
2. Add OPT_COLOR_FLAG(), OPT__COLOR(), and parse_opt_color_flag_cb()
to the option parsing library. The callback uses
git_config_colorbool(), so color.h is now a dependency
of parse-options.c.
Signed-off-by: Mark Lodato <lodatom@gmail.com>
Signed-off-by: Junio C Hamano <gitster@pobox.com>
2010-02-17 12:55:58 +08:00
|
|
|
#define OPT__COLOR(var, h) \
|
|
|
|
OPT_COLOR_FLAG(0, "color", (var), (h))
|
2007-10-14 17:05:12 +08:00
|
|
|
|
2007-10-15 07:35:37 +08:00
|
|
|
#endif
|