Add CallbackRef that takes ref in argument instead of value

[cecton]

Description

It's sometimes useful to pass callback functions to components. The struct
Callback can be used for that but it only accept passing the input argument by
value. It is useful sometimes to pass the input argument by reference instead.

Alternative explanation from @kirillsemyonkin :

Its for doing Fn(&IN) -> OUT instead of Fn(IN) -> OUT since some rust weirdness.

Doing Callback<&IN, OUT> says that the type should have a 'a which is wrong because its Fn under the hood.

It would've worked if we were able to write Callback<&IN> -> OUT or whatever but I think Rust allows that syntax only for Fn family.

Checklist

• I have reviewed my own code
• I have added tests

[github-actions]

Visit the preview URL for this PR (updated for commit 844093c):

https://yew-rs-api--pr3419-callback-ref-eocw38vx.web.app

_{(expires Wed, 04 Oct 2023 12:51:40 GMT)}

_{🔥 via Firebase Hosting GitHub Action 🌎}

[github-actions]

Benchmark - SSR

Yew Master

Benchmark	Round	Min (ms)	Max (ms)	Mean (ms)	Standard Deviation
Baseline	10	319.503	320.053	319.703	0.173
Hello World	10	652.113	668.791	662.738	5.234
Function Router	10	2154.012	2165.300	2159.260	3.924
Concurrent Task	10	1007.025	1008.103	1007.671	0.331
Many Providers	10	1506.672	1533.849	1516.875	8.301

Pull Request

Benchmark	Round	Min (ms)	Max (ms)	Mean (ms)	Standard Deviation
Baseline	10	319.589	319.797	319.696	0.061
Hello World	10	658.760	673.624	668.083	4.246
Function Router	10	2153.638	2169.620	2160.019	5.312
Concurrent Task	10	1006.946	1008.382	1007.784	0.510
Many Providers	10	1529.510	1557.193	1537.271	7.839

[github-actions]

Size Comparison

examples	master (KB)	pull request (KB)	diff (KB)	diff (%)
async_clock	102.960	102.960	0	0.000%
boids	175.618	175.618	0	0.000%
communication_child_to_parent	95.298	95.298	0	0.000%
communication_grandchild_with_grandparent	109.034	109.034	0	0.000%
communication_grandparent_to_grandchild	105.710	105.710	0	0.000%
communication_parent_to_child	92.792	92.792	0	0.000%
contexts	113.453	113.453	0	0.000%
counter	89.196	89.196	0	0.000%
counter_functional	89.936	89.936	0	0.000%
dyn_create_destroy_apps	92.369	92.369	0	0.000%
file_upload	103.513	103.513	0	0.000%
function_memory_game	174.570	174.570	0	0.000%
function_router	353.673	353.673	0	0.000%
function_todomvc	163.505	163.505	0	0.000%
futures	227.445	227.445	0	0.000%
game_of_life	112.163	112.163	0	0.000%
immutable	188.785	188.782	-0.003	-0.002%
inner_html	85.979	85.979	0	0.000%
js_callback	113.458	113.458	0	0.000%
keyed_list	201.203	201.203	0	0.000%
mount_point	89.189	89.189	0	0.000%
nested_list	114.618	114.618	0	0.000%
node_refs	96.290	96.290	0	0.000%
password_strength	1721.062	1721.062	0	0.000%
portals	98.365	98.365	0	0.000%
router	319.654	319.654	0	0.000%
simple_ssr	144.237	144.237	0	0.000%
ssr_router	391.450	391.450	0	0.000%
suspense	119.067	119.067	0	0.000%
timer	91.745	91.745	0	0.000%
timer_functional	100.451	100.451	0	0.000%
todomvc	143.688	143.688	0	0.000%
two_apps	89.897	89.897	0	0.000%
web_worker_fib	138.871	138.871	0	0.000%
web_worker_prime	190.426	190.426	0	0.000%
webgl	88.554	88.554	0	0.000%

✅ None of the examples has changed their size significantly.

[cecton]

I believe this is actually wrong because creating 2 "noop" callbacks and comparing them for equality will yield false

[futursolo]

I think this is fine as it is trying to compare the address of 2 functions are not the content of the functions.

[cecton]

I should have written a test -_-

The PartialEq implementation uses Rc::ptr_eq() so it's actually comparing that it is the same RC but we create one each time we call noop (or Default).

https://play.rust-lang.org/?version=stable&mode=debug&edition=2021&gist=3e6510283a5655f1166410d77fd62d51

[cecton]

I made this PR to address this: #3430

[ranile]

What prevents you from doing Callback<&I, O>

[cecton]

What prevents you from doing Callback<&I, O>

It forces me to provide the lifetime which I have to put in the struct etc... but it's actually not necessary because it's an argument of a function but Rust does not know that

[cecton]

What prevents you from doing Callback<&I, O>

Ah, I forgot to include the important bit from @kirillsemyonkin lol :

doing Callback<&IN, OUT> says that the type should have a 'a which is wrong because its Fn under the hood

[futursolo]

LGTM!