Re: [PATCH] describe: fix --no-exact-match

From: René Scharfe <l.s.r@web.de>
Date: Aug 10, 2023, 19:45 UTC
Message-ID: <b5fcc572-c710-5869-7a40-ba784d8edf0e@web.de>
In-Reply-To: <xmqqo7jfk8xq.fsf@gitster.g>

Am 10.08.23 um 03:00 schrieb Junio C Hamano:

Show 12 quoted lines

> Jeff King <peff@peff.net> writes:
>
>>   int my_foo;
>>   ...
>>   OPT_CALLBACK('f', "foo", int, &my_foo, ...etc);
>>
>> Not great, but it might not be _too_ bad given that most helpers like
>> OPT_BOOL() can just say "int" behind the scenes.
>>
>> That said, I don't recall these void pointers being a large source of
>> errors in the past. So while it's a fun type-system puzzle, the
>> effort/reward ratio might not be favorable.

I have vague memories of changing a callback from using a string_list to a strset (or whatever) and getting crashes out of some other callers that used it non-obviously via some macro. A compile-time check would have helped me.

> I tend to agree on both counts, it is a fun puzzle, and it probably
> is not going to give us sufficient reward.  The fact that "int" and
> typeof(my_foo) need to be manually kept straight defeats the "type
> safety" justification for this mental exercise.

The many-pointers idea is a bit silly, admittedly.

Having to declare a type manually is how most of C's type safety works, though, so a certain amount of that is unavoidable.

typeof (along with typeof_unqual) is nice, but I guess it will take a few years (or decades?) before we can expect it to be supported by all relevant platforms.

The typed-callback-with-void-wrapper idea seems promising. Let's see if we can get it into usable shape.

René