This macro doesn't work and i don't know why. I've tried multiple
different things to try to delay the expansion of _0MAKE_LISTENER and none
of them have worked. I don't know if there's a more idiomatic way to do
this. I've tried looking at the output from putting it through ‘cpp’ and
that didn't work because it errored out before i could see what it
expanded to.

#define MAKE_LISTENER(object_type, ...) static const struct\
object_type##_listener object_type##_listener = { P99_SEQ(\
_0MAKE_LISTENER, P99_SEQ((object_type), __VA_ARGS__)) }
#define _0MAKE_LISTENER(x) _1MAKE_LISTENER (x, _2MAKE_LISTENER
#define _1MAKE_LISTENER _3MAKE_LISTENER
#define _2MAKE_LISTENER(y) y)
#define _3MAKE_LISTENER _4MAKE_LISTENER
#define _4MAKE_LISTENER(object_type, event_name) .event_name =\
callback_##object_type##_##event_name

--
Blue-Maned_Hawk│shortens to Hawk│/blu.mɛin.dʰak/│he/him/his/himself/Mr.
blue-maned_hawk.srht.site
Like many 40-year-olds, these machines have more health issues than
expected.

Blue-Maned_Hawk <bluemanedhawk@invalid.invalid> writes:

> This macro doesn't work and i don't know why. I've tried multiple
> different things to try to delay the expansion of _0MAKE_LISTENER and none
> of them have worked. I don't know if there's a more idiomatic way to do
> this. I've tried looking at the output from putting it through ?cpp? and
> that didn't work because it errored out before i could see what it
> expanded to.
>
> #define MAKE_LISTENER(object_type, ...) static const struct\
> object_type##_listener object_type##_listener = { P99_SEQ(\
> _0MAKE_LISTENER, P99_SEQ((object_type), __VA_ARGS__)) }
> #define _0MAKE_LISTENER(x) _1MAKE_LISTENER (x, _2MAKE_LISTENER
> #define _1MAKE_LISTENER _3MAKE_LISTENER
> #define _2MAKE_LISTENER(y) y)
> #define _3MAKE_LISTENER _4MAKE_LISTENER
> #define _4MAKE_LISTENER(object_type, event_name) .event_name =\
> callback_##object_type##_##event_name

What is it that you want to accomplish?

Tim Rentsch wrote:

> What is it that you want to accomplish?

As i stated in the subject line, i am trying to create a macro to create
Wayland listeners.

--
Blue-Maned_Hawk│shortens to Hawk│/blu.mɛin.dʰak/│he/him/his/himself/Mr.
blue-maned_hawk.srht.site
Nothing's thoroughly disposed of until it's been through a digestive
system.

On 09/04/2024 20:17, Blue-Maned_Hawk wrote:
> Tim Rentsch wrote:
>
>> What is it that you want to accomplish?
>
> As i stated in the subject line, i am trying to create a macro to create
> Wayland listeners.
>

Assume, for the sake of argument, that no one knows or cares anything
about Wayland or listeners.

You are trying to use a set macros to expand to some code. Start by
showing what code you want to get from examples of using your macros -
then people can try to help get the macros right.

David Brown wrote:

> On 09/04/2024 20:17, Blue-Maned_Hawk wrote:
>> Tim Rentsch wrote:
>>
>>> What is it that you want to accomplish?
>>
>> As i stated in the subject line, i am trying to create a macro to
>> create Wayland listeners.
>>
>>
> Assume, for the sake of argument, that no one knows or cares anything
> about Wayland or listeners.

<https://xkcd.com/1432/>

> You are trying to use a set macros to expand to some code. Start by
> showing what code you want to get from examples of using your macros -
> then people can try to help get the macros right.

MAKE_LISTENER(wl_registry, global, global_remove);

→

static const struct wl_registry_listener wl_registry_listener = { .global
= callback_wl_registry_global, .global_remove =
callback_wl_registry_global_remove };

–

MAKE_LISTENER(xdg_wm_base, configure, close, configure_bounds,
wm_capabilities);

→

static const struct xdg_wm_base_listener xdg_wm_base_listener =
{ .configure = callback_xdg_wm_base_configure, .close =
callback_xdg_wm_base_close, .configure_bounds =
callback_xdg_wm_base_configure_bounds, .wm_capabilities =
callback_xdg_wm_base_wm_capabilities };

--
Blue-Maned_Hawk│shortens to Hawk│/blu.mɛin.dʰak/│he/him/his/himself/Mr.
blue-maned_hawk.srht.site
Eating carpets strictly prohibited!

On 4/8/24 20:36, Blue-Maned_Hawk wrote:
> This macro doesn't work and i don't know why. I've tried multiple
> different things to try to delay the expansion of _0MAKE_LISTENER and none
> of them have worked. I don't know if there's a more idiomatic way to do
> this. I've tried looking at the output from putting it through ‘cpp’ and
> that didn't work because it errored out before i could see what it
> expanded to.
>
> #define MAKE_LISTENER(object_type, ...) static const struct\
> object_type##_listener object_type##_listener = { P99_SEQ(\

This isn't a C problem, or at least,not entirely a C problem. I'd
thought that the P99 in your subject line might be a typo for C99, but I
have no idea what P99_SEQ is. Wikipedia knows 9 different meanings for
P99. Google gives me way too many hits from P99, I have no idea which
one is relevant. You might need to take this up with someone who has
some idea what P99 is.

The basic technique for delaying expansion is as follows:

#define STR(e) #e
#define STRSTR(e) STR(e)

and then replace WHATEVER with STRSTR(WHATEVER).

There are, I understand, good reasons based upon how macro expansion
works why this solution is necessary, but even when I was a lot younger
I had trouble understanding those reasons. I generally had to rely upon
trial and error to figure out the right way to use STRSTR().

On 09/04/2024 22:20, Blue-Maned_Hawk wrote:
> David Brown wrote:
>
>> On 09/04/2024 20:17, Blue-Maned_Hawk wrote:
>>> Tim Rentsch wrote:
>>>
>>>> What is it that you want to accomplish?
>>>
>>> As i stated in the subject line, i am trying to create a macro to
>>> create Wayland listeners.
>>>
>>>
>> Assume, for the sake of argument, that no one knows or cares anything
>> about Wayland or listeners.
>
> <https://xkcd.com/1432/>
>
>> You are trying to use a set macros to expand to some code. Start by
>> showing what code you want to get from examples of using your macros -
>> then people can try to help get the macros right.
>
> MAKE_LISTENER(wl_registry, global, global_remove);
>
> →
>
> static const struct wl_registry_listener wl_registry_listener = { .global
> = callback_wl_registry_global, .global_remove =
> callback_wl_registry_global_remove };

#define _get_6th_arg(_1, _2, _3, _4, _5, _arg, ...) _arg
#define _fe_0(_global, _call, ...)
#define _fe_1(_global, _call, x) _call(_global, x)
#define _fe_2(_global, _call, x, ...) _call(_global, x) _fe_1(_global,
_call, __VA_ARGS__)
#define _fe_3(_global, _call, x, ...) _call(_global, x) _fe_2(_global,
_call, __VA_ARGS__)
#define _fe_4(_global, _call, x, ...) _call(_global, x) _fe_3(_global,
_call, __VA_ARGS__)
#define expand_macro_for_each(_global, _macro, ...) \
_get_6th_arg("ignored", __VA_ARGS__, \
_fe_4, _fe_3, _fe_2, _fe_1, _fe_0)(_global, _macro, __VA_ARGS__)

#define MAKE_LISTENER(object_type, ...) _make_listener(object_type,
__VA_ARGS__)
#define _listener_entry(_global, _field) , . _field = callback_ ##
_global ## _ ## field
#define _make_listener(object_type, first_field, ...) static const struct \
object_type ## _listener object_type ## _listener = { \
.first_field = callback_ ## object_type ## _ ## first_field \
expand_macro_for_each(object_type, _listener_entry, __VA_ARGS__) \
}

James Kuyper wrote:

> This isn't a C problem, or at least,not entirely a C problem. I'd
> thought that the P99 in your subject line might be a typo for C99, but I
> have no idea what P99_SEQ is. Wikipedia knows 9 different meanings for
> P99. Google gives me way too many hits from P99, I have no idea which
> one is relevant. You might need to take this up with someone who has
> some idea what P99 is.

This is the P99 that i am referring to: <https://
gustedt.gitlabpages.inria.fr/p99/p99-html/index.html>.

--
Blue-Maned_Hawk│shortens to Hawk│/blu.mɛin.dʰak/│he/him/his/himself/Mr.
blue-maned_hawk.srht.site
Indisputably a ring!

David Brown <david.brown@hesbynett.no> writes:
>On 09/04/2024 22:20, Blue-Maned_Hawk wrote:
>> David Brown wrote:
>>
>>> On 09/04/2024 20:17, Blue-Maned_Hawk wrote:
>>>> Tim Rentsch wrote:
>>>>
>>>>> What is it that you want to accomplish?
>>>>
>>>> As i stated in the subject line, i am trying to create a macro to
>>>> create Wayland listeners.
>>>>
>>>>
>>> Assume, for the sake of argument, that no one knows or cares anything
>>> about Wayland or listeners.
>>
>> <https://xkcd.com/1432/>
>>
>>> You are trying to use a set macros to expand to some code. Start by
>>> showing what code you want to get from examples of using your macros -
>>> then people can try to help get the macros right.
>>
>> MAKE_LISTENER(wl_registry, global, global_remove);
>>
>> →
>>
>> static const struct wl_registry_listener wl_registry_listener = { .global
>> = callback_wl_registry_global, .global_remove =
>> callback_wl_registry_global_remove };
>
>#define _get_6th_arg(_1, _2, _3, _4, _5, _arg, ...) _arg
>#define _fe_0(_global, _call, ...)
>#define _fe_1(_global, _call, x) _call(_global, x)
>#define _fe_2(_global, _call, x, ...) _call(_global, x) _fe_1(_global,
>_call, __VA_ARGS__)
>#define _fe_3(_global, _call, x, ...) _call(_global, x) _fe_2(_global,
>_call, __VA_ARGS__)
>#define _fe_4(_global, _call, x, ...) _call(_global, x) _fe_3(_global,
>_call, __VA_ARGS__)
>#define expand_macro_for_each(_global, _macro, ...) \
> _get_6th_arg("ignored", __VA_ARGS__, \
> _fe_4, _fe_3, _fe_2, _fe_1, _fe_0)(_global, _macro, __VA_ARGS__)
>
>
>
>#define MAKE_LISTENER(object_type, ...) _make_listener(object_type,
>__VA_ARGS__)
>#define _listener_entry(_global, _field) , . _field = callback_ ##
>_global ## _ ## field
>#define _make_listener(object_type, first_field, ...) static const struct \
> object_type ## _listener object_type ## _listener = { \
> .first_field = callback_ ## object_type ## _ ## first_field \
> expand_macro_for_each(object_type, _listener_entry, __VA_ARGS__) \
> }
>

Frankly I'd reject that code in any code review request.

David Brown <david.brown@hesbynett.no> writes:

> On 09/04/2024 22:20, Blue-Maned_Hawk wrote:
>> David Brown wrote:
>>
>>> On 09/04/2024 20:17, Blue-Maned_Hawk wrote:
>>>> Tim Rentsch wrote:
>>>>
>>>>> What is it that you want to accomplish?
>>>>
>>>> As i stated in the subject line, i am trying to create a macro to
>>>> create Wayland listeners.
>>>>
>>>>
>>> Assume, for the sake of argument, that no one knows or cares anything
>>> about Wayland or listeners.
>> <https://xkcd.com/1432/>
>>
>>> You are trying to use a set macros to expand to some code. Start by
>>> showing what code you want to get from examples of using your macros -
>>> then people can try to help get the macros right.
>> MAKE_LISTENER(wl_registry, global, global_remove);
>> →
>> static const struct wl_registry_listener wl_registry_listener = { .global
>> = callback_wl_registry_global, .global_remove =
>> callback_wl_registry_global_remove };
>
> #define _get_6th_arg(_1, _2, _3, _4, _5, _arg, ...) _arg
> #define _fe_0(_global, _call, ...)
> #define _fe_1(_global, _call, x) _call(_global, x)
> #define _fe_2(_global, _call, x, ...) _call(_global, x) _fe_1(_global,
> _call, __VA_ARGS__)
> #define _fe_3(_global, _call, x, ...) _call(_global, x) _fe_2(_global,
> _call, __VA_ARGS__)
> #define _fe_4(_global, _call, x, ...) _call(_global, x) _fe_3(_global,
> _call, __VA_ARGS__)
> #define expand_macro_for_each(_global, _macro, ...) \
> _get_6th_arg("ignored", __VA_ARGS__, \
> _fe_4, _fe_3, _fe_2, _fe_1, _fe_0)(_global, _macro, __VA_ARGS__)
>
>
> #define MAKE_LISTENER(object_type, ...) _make_listener(object_type,
> __VA_ARGS__)
> #define _listener_entry(_global, _field) , . _field = callback_ ## _global
> ## _ ## field

s/field/_field/

> #define _make_listener(object_type, first_field, ...) static const struct \
> object_type ## _listener object_type ## _listener = { \
> .first_field = callback_ ## object_type ## _ ## first_field \
> expand_macro_for_each(object_type, _listener_entry, __VA_ARGS__) \
> }

--
Ben.

On 10/04/2024 16:26, Scott Lurndal wrote:
> David Brown <david.brown@hesbynett.no> writes:
>> On 09/04/2024 22:20, Blue-Maned_Hawk wrote:
>>> David Brown wrote:
>>>
>>>> On 09/04/2024 20:17, Blue-Maned_Hawk wrote:
>>>>> Tim Rentsch wrote:
>>>>>
>>>>>> What is it that you want to accomplish?
>>>>>
>>>>> As i stated in the subject line, i am trying to create a macro to
>>>>> create Wayland listeners.
>>>>>
>>>>>
>>>> Assume, for the sake of argument, that no one knows or cares anything
>>>> about Wayland or listeners.
>>>
>>> <https://xkcd.com/1432/>
>>>
>>>> You are trying to use a set macros to expand to some code. Start by
>>>> showing what code you want to get from examples of using your macros -
>>>> then people can try to help get the macros right.
>>>
>>> MAKE_LISTENER(wl_registry, global, global_remove);
>>>
>>> →
>>>
>>> static const struct wl_registry_listener wl_registry_listener = { .global
>>> = callback_wl_registry_global, .global_remove =
>>> callback_wl_registry_global_remove };
>>
>> #define _get_6th_arg(_1, _2, _3, _4, _5, _arg, ...) _arg
>> #define _fe_0(_global, _call, ...)
>> #define _fe_1(_global, _call, x) _call(_global, x)
>> #define _fe_2(_global, _call, x, ...) _call(_global, x) _fe_1(_global,
>> _call, __VA_ARGS__)
>> #define _fe_3(_global, _call, x, ...) _call(_global, x) _fe_2(_global,
>> _call, __VA_ARGS__)
>> #define _fe_4(_global, _call, x, ...) _call(_global, x) _fe_3(_global,
>> _call, __VA_ARGS__)
>> #define expand_macro_for_each(_global, _macro, ...) \
>> _get_6th_arg("ignored", __VA_ARGS__, \
>> _fe_4, _fe_3, _fe_2, _fe_1, _fe_0)(_global, _macro, __VA_ARGS__)
>>
>>
>>
>> #define MAKE_LISTENER(object_type, ...) _make_listener(object_type,
>> __VA_ARGS__)
>> #define _listener_entry(_global, _field) , . _field = callback_ ##
>> _global ## _ ## field
>> #define _make_listener(object_type, first_field, ...) static const struct \
>> object_type ## _listener object_type ## _listener = { \
>> .first_field = callback_ ## object_type ## _ ## first_field \
>> expand_macro_for_each(object_type, _listener_entry, __VA_ARGS__) \
>> }
>>
>
> Frankly I'd reject that code in any code review request.

I normally like code to be so clear that comments about how it works are
unnecessary. But I think for something like this, a few comments would
be helpful!

When doing this kind of messing around with advanced preprocessing, it's
useful to have a set of re-usable macros that can be written and
documented once. The "expand_macro_for_each" macro and its parts is
partly there, and once that is in place and accepted, the rest is a lot
smaller step to accept in a code review.

But this is code to help the OP get his macro generations working.
Whether he likes it, rejects it, or uses it for inspiration is up to him.

Blue-Maned_Hawk <bluemanedhawk@invalid.invalid> writes:

[request for example]

> MAKE_LISTENER(wl_registry, global, global_remove);
>
> ?
>
> static const struct wl_registry_listener wl_registry_listener = {
> .global = callback_wl_registry_global,
> .global_remove = callback_wl_registry_global_remove
> };
>
> ?
>
> MAKE_LISTENER(xdg_wm_base, configure, close, configure_bounds,
> wm_capabilities);
>
> ?
>
> static const struct xdg_wm_base_listener xdg_wm_base_listener = {
> .configure = callback_xdg_wm_base_configure,
> .close = callback_xdg_wm_base_close,
> .configure_bounds = callback_xdg_wm_base_configure_bounds,
> .wm_capabilities = callback_xdg_wm_base_wm_capabilities
> };

I think it isn't too hard to write a macro to do what you
want. I'm pretty sure it's possible to define such a macro
using the P99 macros to do the iteration, but I don't see
an easy way to do it. If I were in need of such a macro,
probably I would ignore the P99 macros and just build a
simpler solution that works for the task at hand.

James Kuyper <jameskuyper@alumni.caltech.edu> writes:

> On 4/8/24 20:36, Blue-Maned_Hawk wrote:
>
>> This macro doesn't work and i don't know why. I've tried multiple
>> different things to try to delay the expansion of _0MAKE_LISTENER and none
>> of them have worked. I don't know if there's a more idiomatic way to do
>> this. I've tried looking at the output from putting it through ?cpp? and
>> that didn't work because it errored out before i could see what it
>> expanded to.
>>
>> #define MAKE_LISTENER(object_type, ...) static const struct\
>> object_type##_listener object_type##_listener = { P99_SEQ(\
>
> This isn't a C problem, or at least,not entirely a C problem. I'd
> thought that the P99 in your subject line might be a typo for C99, but I
> have no idea what P99_SEQ is. Wikipedia knows 9 different meanings for
> P99. [...]

Someone might tell James that it is entirely a C problem for
anyone with enough sense to search for P99_SEQ rather than
just P99.

Ben Bacarisse <ben.usenet@bsb.me.uk> writes:

> David Brown <david.brown@hesbynett.no> writes:
>
>> On 09/04/2024 22:20, Blue-Maned_Hawk wrote:
>>
>>> David Brown wrote:
>>>
>>>> On 09/04/2024 20:17, Blue-Maned_Hawk wrote:
>>>>
>>>>> Tim Rentsch wrote:
>>>>>
>>>>>> What is it that you want to accomplish?
>>>>>
>>>>> As i stated in the subject line, i am trying to create a macro to
>>>>> create Wayland listeners.
>>>>
>>>> Assume, for the sake of argument, that no one knows or cares anything
>>>> about Wayland or listeners.
>>>
>>> <https://xkcd.com/1432/>
>>>
>>>> You are trying to use a set macros to expand to some code. Start by
>>>> showing what code you want to get from examples of using your macros -
>>>> then people can try to help get the macros right.
>>>
>>> MAKE_LISTENER(wl_registry, global, global_remove);
>>> ?
>>> static const struct wl_registry_listener wl_registry_listener = { .global
>>> = callback_wl_registry_global, .global_remove =
>>> callback_wl_registry_global_remove };
>>
>> #define _get_6th_arg(_1, _2, _3, _4, _5, _arg, ...) _arg
>> #define _fe_0(_global, _call, ...)
>> #define _fe_1(_global, _call, x) _call(_global, x)
>> #define _fe_2(_global, _call, x, ...) _call(_global, x) _fe_1(_global,
>> _call, __VA_ARGS__)
>> #define _fe_3(_global, _call, x, ...) _call(_global, x) _fe_2(_global,
>> _call, __VA_ARGS__)
>> #define _fe_4(_global, _call, x, ...) _call(_global, x) _fe_3(_global,
>> _call, __VA_ARGS__)
>> #define expand_macro_for_each(_global, _macro, ...) \
>> _get_6th_arg("ignored", __VA_ARGS__, \
>> _fe_4, _fe_3, _fe_2, _fe_1, _fe_0)(_global, _macro, __VA_ARGS__)
>>
>>
>> #define MAKE_LISTENER(object_type, ...) _make_listener(object_type,
>> __VA_ARGS__)
>> #define _listener_entry(_global, _field) , . _field = callback_ ## _global
>> ## _ ## field
>
> s/field/_field/

The bigger problem is that the proposed solution doesn't work
for more than five fields. Clearly the idea is to define a
macro that works for an unlimited number of fields, or at least
a fairly large number.

On 10/04/2024 18:08, Tim Rentsch wrote:
> Ben Bacarisse <ben.usenet@bsb.me.uk> writes:
>
>> David Brown <david.brown@hesbynett.no> writes:
>>
>>> On 09/04/2024 22:20, Blue-Maned_Hawk wrote:
>>>
>>>> David Brown wrote:
>>>>
>>>>> On 09/04/2024 20:17, Blue-Maned_Hawk wrote:
>>>>>
>>>>>> Tim Rentsch wrote:
>>>>>>
>>>>>>> What is it that you want to accomplish?
>>>>>>
>>>>>> As i stated in the subject line, i am trying to create a macro to
>>>>>> create Wayland listeners.
>>>>>
>>>>> Assume, for the sake of argument, that no one knows or cares anything
>>>>> about Wayland or listeners.
>>>>
>>>> <https://xkcd.com/1432/>
>>>>
>>>>> You are trying to use a set macros to expand to some code. Start by
>>>>> showing what code you want to get from examples of using your macros -
>>>>> then people can try to help get the macros right.
>>>>
>>>> MAKE_LISTENER(wl_registry, global, global_remove);
>>>> ?
>>>> static const struct wl_registry_listener wl_registry_listener = { .global
>>>> = callback_wl_registry_global, .global_remove =
>>>> callback_wl_registry_global_remove };
>>>
>>> #define _get_6th_arg(_1, _2, _3, _4, _5, _arg, ...) _arg
>>> #define _fe_0(_global, _call, ...)
>>> #define _fe_1(_global, _call, x) _call(_global, x)
>>> #define _fe_2(_global, _call, x, ...) _call(_global, x) _fe_1(_global,
>>> _call, __VA_ARGS__)
>>> #define _fe_3(_global, _call, x, ...) _call(_global, x) _fe_2(_global,
>>> _call, __VA_ARGS__)
>>> #define _fe_4(_global, _call, x, ...) _call(_global, x) _fe_3(_global,
>>> _call, __VA_ARGS__)
>>> #define expand_macro_for_each(_global, _macro, ...) \
>>> _get_6th_arg("ignored", __VA_ARGS__, \
>>> _fe_4, _fe_3, _fe_2, _fe_1, _fe_0)(_global, _macro, __VA_ARGS__)
>>>
>>>
>>> #define MAKE_LISTENER(object_type, ...) _make_listener(object_type,
>>> __VA_ARGS__)
>>> #define _listener_entry(_global, _field) , . _field = callback_ ## _global
>>> ## _ ## field
>>
>> s/field/_field/

Correct - sorry about the typo.

>
> The bigger problem is that the proposed solution doesn't work
> for more than five fields.

I have no idea how many fields the OP might want, in order to judge if
that is a big problem or not. But I believe he'll be able to figure out
how to extend it as necessary.

> Clearly the idea is to define a
> macro that works for an unlimited number of fields, or at least
> a fairly large number.

If you have a way to extend it to unlimited numbers of fields, I would
love to know - as far as I am aware, the lack of recursion or loops in
the C pre-processor means that you always have to be limited like this.
It is, I think, possible to make sequence of macros like this that split
lists in half, so that you only need O(log N) submacros in order to
support up to N arguments. That would let you get to the level of very
large numbers pretty quickly.

On 4/10/2024 7:54 AM, David Brown wrote:
> On 10/04/2024 16:26, Scott Lurndal wrote:
>> David Brown <david.brown@hesbynett.no> writes:
>>> On 09/04/2024 22:20, Blue-Maned_Hawk wrote:
>>>> David Brown wrote:
>>>>
>>>>> On 09/04/2024 20:17, Blue-Maned_Hawk wrote:
>>>>>> Tim Rentsch wrote:
>>>>>>
>>>>>>> What is it that you want to accomplish?
>>>>>>
>>>>>> As i stated in the subject line, i am trying to create a macro to
>>>>>> create Wayland listeners.
>>>>>>
>>>>>>
>>>>> Assume, for the sake of argument, that no one knows or cares anything
>>>>> about Wayland or listeners.
>>>>
>>>> <https://xkcd.com/1432/>
>>>>
>>>>> You are trying to use a set macros to expand to some code.  Start by
>>>>> showing what code you want to get from examples of using your macros -
>>>>> then people can try to help get the macros right.
>>>>
>>>> MAKE_LISTENER(wl_registry, global, global_remove);
>>>>
>>>> →
>>>>
>>>> static const struct wl_registry_listener wl_registry_listener = {
>>>> .global
>>>> = callback_wl_registry_global, .global_remove =
>>>> callback_wl_registry_global_remove };
>>>
>>> #define _get_6th_arg(_1, _2, _3, _4, _5, _arg, ...) _arg
>>> #define _fe_0(_global, _call, ...)
>>> #define _fe_1(_global, _call, x) _call(_global, x)
>>> #define _fe_2(_global, _call, x, ...) _call(_global, x) _fe_1(_global,
>>> _call, __VA_ARGS__)
>>> #define _fe_3(_global, _call, x, ...) _call(_global, x) _fe_2(_global,
>>> _call, __VA_ARGS__)
>>> #define _fe_4(_global, _call, x, ...) _call(_global, x) _fe_3(_global,
>>> _call, __VA_ARGS__)
>>> #define expand_macro_for_each(_global, _macro, ...) \
>>>      _get_6th_arg("ignored", __VA_ARGS__, \
>>>      _fe_4, _fe_3, _fe_2, _fe_1, _fe_0)(_global, _macro, __VA_ARGS__)
>>>
>>>
>>>
>>> #define MAKE_LISTENER(object_type, ...) _make_listener(object_type,
>>> __VA_ARGS__)
>>> #define _listener_entry(_global, _field) , . _field = callback_ ##
>>> _global ## _ ## field
>>> #define _make_listener(object_type, first_field, ...) static const
>>> struct \
>>>      object_type ## _listener object_type ## _listener = { \
>>>      .first_field = callback_ ## object_type ## _ ## first_field \
>>>      expand_macro_for_each(object_type, _listener_entry, __VA_ARGS__) \
>>>      }
>>>
>>
>> Frankly I'd reject that code in any code review request.
>
> I normally like code to be so clear that comments about how it works are
> unnecessary.  But I think for something like this, a few comments would
> be helpful!
>
> When doing this kind of messing around with advanced preprocessing, it's
> useful to have a set of re-usable macros that can be written and
> documented once.  The "expand_macro_for_each" macro and its parts is
> partly there, and once that is in place and accepted, the rest is a lot
> smaller step to accept in a code review.
>
> But this is code to help the OP get his macro generations working.
> Whether he likes it, rejects it, or uses it for inspiration is up to him.
>
>

The OP needs a deep read of the chaos lib:

https://github.com/rofl0r/chaos-pp

Might help... ? ;^)

On 10/04/2024 20:44, Chris M. Thomasson wrote:

>
> The OP needs a deep read of the chaos lib:
>
> https://github.com/rofl0r/chaos-pp
>
> Might help... ? ;^)

I think more people would respond to your posts if you actually included
content, rather than just random links and smilies.

Assume that no one here will ever click on a youtube link - then you
clearly don't need to post them.

Assume that no one will click on any random webpage link, or even to
known sites like github, unless there's plenty of information in the
posts to make it clear why the linked software is relevant.

You do a lot of interesting programming, have interesting ideas, and are
energetic about sharing them and helping people. That's all great. But
you could do so much better at getting your thoughts across.

So if you don't mind - please try again. Tell us what "the chaos lib"
is. Say why it is exciting. Explain how it might help the OP, or
anyone else.

On 10/04/2024 20:29, David Brown wrote:
> On 10/04/2024 20:44, Chris M. Thomasson wrote:
>
>>
>> The OP needs a deep read of the chaos lib:
>>
>> https://github.com/rofl0r/chaos-pp
>>
>> Might help... ? ;^)
>
> I think more people would respond to your posts if you actually included
> content, rather than just random links and smilies.
>
> Assume that no one here will ever click on a youtube link - then you
> clearly don't need to post them.
>
> Assume that no one will click on any random webpage link, or even to
> known sites like github, unless there's plenty of information in the
> posts to make it clear why the linked software is relevant.
>
> You do a lot of interesting programming, have interesting ideas, and are
> energetic about sharing them and helping people.  That's all great.  But
> you could do so much better at getting your thoughts across.
>
> So if you don't mind - please try again.  Tell us what "the chaos lib"
> is.  Say why it is exciting.  Explain how it might help the OP, or
> anyone else.
>

I spent a few minutes looking at the link. I didn't learn anything. Just
loads of raw XML, XSL and HTML files. No explanation anywhere of what is
what.

The only two files in plaintext were LICENCE and README, both showing
licence info.

I've seen lots of github projects and this is poor. But then, I don't
know if this was ever meant to be seen like this; maybe it is just data
referenced by another project. Or you need special tools to view.

On 4/10/2024 12:29 PM, David Brown wrote:
> On 10/04/2024 20:44, Chris M. Thomasson wrote:
>
>>
>> The OP needs a deep read of the chaos lib:
>>
>> https://github.com/rofl0r/chaos-pp
>>
>> Might help... ? ;^)
>
> I think more people would respond to your posts if you actually included
> content, rather than just random links and smilies.
>
> Assume that no one here will ever click on a youtube link - then you
> clearly don't need to post them.
>
> Assume that no one will click on any random webpage link, or even to
> known sites like github, unless there's plenty of information in the
> posts to make it clear why the linked software is relevant.
>
> You do a lot of interesting programming, have interesting ideas, and are
> energetic about sharing them and helping people.  That's all great.  But
> you could do so much better at getting your thoughts across.
>
> So if you don't mind - please try again.  Tell us what "the chaos lib"
> is.  Say why it is exciting.  Explain how it might help the OP, or
> anyone else.
>

It is a preprocessor library with a lot of interesting fancy macros.
Iirc, the chaos preprocessor library is used in Boost.

On 4/10/2024 1:40 PM, Chris M. Thomasson wrote:
> On 4/10/2024 12:29 PM, David Brown wrote:
>> On 10/04/2024 20:44, Chris M. Thomasson wrote:
>>
>>>
>>> The OP needs a deep read of the chaos lib:
>>>
>>> https://github.com/rofl0r/chaos-pp
>>>
>>> Might help... ? ;^)
>>
>> I think more people would respond to your posts if you actually
>> included content, rather than just random links and smilies.
>>
>> Assume that no one here will ever click on a youtube link - then you
>> clearly don't need to post them.
>>
>> Assume that no one will click on any random webpage link, or even to
>> known sites like github, unless there's plenty of information in the
>> posts to make it clear why the linked software is relevant.
>>
>> You do a lot of interesting programming, have interesting ideas, and
>> are energetic about sharing them and helping people.  That's all
>> great.  But you could do so much better at getting your thoughts across.
>>
>> So if you don't mind - please try again.  Tell us what "the chaos lib"
>> is.  Say why it is exciting.  Explain how it might help the OP, or
>> anyone else.
>>
>
> It is a preprocessor library with a lot of interesting fancy macros.
> Iirc, the chaos preprocessor library is used in Boost.

Yeah. Sigh. However, you can read some of the docs. A screenshot from my
end dealing with preprocessor arrays:

https://i.ibb.co/1rhBjCf/image.png

Another one dealing with so-called "active arguments":

https://i.ibb.co/vPTvdnP/image.png

I think Boost used to use or is based on chaos-pp wrt BOOST_PP_*

https://i.ibb.co/vPTvdnP/image.png

On 4/10/2024 12:39 PM, bart wrote:
> On 10/04/2024 20:29, David Brown wrote:
>> On 10/04/2024 20:44, Chris M. Thomasson wrote:
>>
>>>
>>> The OP needs a deep read of the chaos lib:
>>>
>>> https://github.com/rofl0r/chaos-pp
[...]
> I've seen lots of github projects and this is poor. But then, I don't
> know if this was ever meant to be seen like this; maybe it is just data
> referenced by another project. Or you need special tools to view.

I accidentally posted this followup to David Brown. Sorry!

Yeah. Sigh. However, you can read some of the docs. A screenshot from my
end dealing with preprocessor arrays:

https://i.ibb.co/1rhBjCf/image.png

Another one dealing with so-called "active arguments":

https://i.ibb.co/vPTvdnP/image.png

I think Boost used to use or is based on chaos-pp wrt BOOST_PP_*

https://i.ibb.co/vPTvdnP/image.png

Also, remember this old post Bart?

https://groups.google.com/g/comp.lang.c/c/LkaB9kxIzTE/m/VypqT2VRBgAJ

:^)

On Wed, 10 Apr 2024 02:49:03 -0400, James Kuyper wrote:

> There are, I understand, good reasons based upon how macro expansion
> works why this solution is necessary, but even when I was a lot younger
> I had trouble understanding those reasons.

String-based macros are like that. That’s why Lisp-style macros are a
better idea.

Blue-Maned_Hawk <bluemanedhawk@invalid.invalid> writes:

> This macro doesn't work and i don't know why. I've tried multiple
> different things to try to delay the expansion of _0MAKE_LISTENER
> and none of them have worked. I don't know if there's a more
> idiomatic way to do this. I've tried looking at the output from
> putting it through ?cpp? and that didn't work because it errored
> out before i could see what it expanded to.
>
> #define MAKE_LISTENER(object_type, ...) static const struct\
> object_type##_listener object_type##_listener = { P99_SEQ(\
> _0MAKE_LISTENER, P99_SEQ((object_type), __VA_ARGS__)) }
> #define _0MAKE_LISTENER(x) _1MAKE_LISTENER (x, _2MAKE_LISTENER
> #define _1MAKE_LISTENER _3MAKE_LISTENER
> #define _2MAKE_LISTENER(y) y)
> #define _3MAKE_LISTENER _4MAKE_LISTENER
> #define _4MAKE_LISTENER(object_type, event_name) .event_name =\
> callback_##object_type##_##event_name

After a lot of trial and error, I came up with the code below.
Probably it could be simplified, but I was happy just to get
it to work.

#include "p99.h"

#define static_const_LISTENER( kind, ... ) \
static const PLAIN_LISTENER( kind, __VA_ARGS__ )

#define PLAIN_LISTENER( kind, ... ) \
struct LISTENER_TAG_NAME( kind ) LISTENER_TAG_NAME( kind ) = { \
LISTENER_FIELDS( kind, __VA_ARGS__ ) \
}

#define LISTENER_TAG_NAME( kind ) JOIN2( kind, _listener )

#define LISTENER_FIELDS( kind, ... ) \
P99_SEQ( AS_IS, \
P99_SEQ( ONE_LISTENER_FIELD_2 AS_IS, \
P99_SEQ( ONE_LISTENER_FIELD, P99_SEQ( (kind), __VA_ARGS__ ) ) \
) \
)

#define ONE_LISTENER_FIELD( a ) ONE_LISTENER_FIELDx( a )
#define ONE_LISTENER_FIELDx( a ) SINGLES_INTO_PAIR( a )

#define SINGLES_INTO_PAIR( a_b ) ( ONE_COMMA_TWO a_b )
#define ONE_COMMA_TWO( a ) a, AS_IS

#define AS_IS( a ) a

#define ONE_LISTENER_FIELD_2( a, b ) ONE_LISTENER_FIELD_2x( a, b )
#define ONE_LISTENER_FIELD_2x( kind, field ) \
.field = callback_ ## kind ## _ ## field

#define JOIN2( a, b ) JOIN2x( a, b )
#define JOIN2x( a, b ) a ## b

static_const_LISTENER( wl_registry, global, global_remove );
static_const_LISTENER(
xdg_wm_base,
configure, close, configure_bounds, wm_capabilities
);