get_just_pressed and get_just_released

[ScriptLineStudios]

duplicate of my old pr - pygame/pygame#3237

[oddbookworm]

I'd be happy to approve this once you update the versionadded tags to 2.1.4

[robertpfeiffer]

I really want this functionality merged, because it's something that exists in HaxeFlixel and Unity and everywhere. My main concern is what API/behaviour/guarantees this should give. I think as is, it wouldn't work in any of my games that get() different event types. I think the reset should only be triggered when dopump is true, or better yet, the reset code should live inside _pg_event_pump. You have my approval conditional on these changes.

I know this change will make the docs slightly more unwieldy, and I don't know how your code will interact with keyboard events introduced by pygame.event.post, but I think this is worth including now.

[MyreMylar]

Updating the version strings

[oddbookworm]

LGTM!

[robertpfeiffer]

I don't think resetting on pygame.event.get() only is a good idea, and resetting on pygame.event.get(pump=False) is a very bad idea. I think it should reset on calls to pump.

I like the idea in general, I like the API, but I think we need to talk about this a little more. I'd also be open to implementing the change myself.

[robertpfeiffer]

Say the word and I push the changes I'd like to your branch

[Starbuck5]

Directly pushed fixes for couple small issues I saw with it.

I like your changes @robertpfeiffer, this seems like a good use for an event filter.

This has got to be a record for the most people who've committed to a single PR branch?

[Starbuck5]

@ankith26 Thanks for taking a look at this.

About your reviews on this PR, you should push those changes directly. Scriptline seems busy right now, and more focused on their recent PRs. Plus, we've got a record to set with the committers on this branch!

[MyreMylar]

This has developed a small conflict. Looks like it could be ready to go for 2.4.0 with a tiny bit of love as it already has three approvals.

[ankith26]

Some reviews (mostly notes to self) that I will address when I get back to this PR later today

[ankith26]

I have pushed some changes, and now this PR looks good to go for me.

I also wrote a little test script to validate that my changes don't break anything, here is it for anyone else interested in testing things out

import pygame

win = pygame.display.set_mode((500, 500))

letters = [pygame.K_a + i for i in range(26)]

while True:
    cur_pressed = set()
    cur_released = set()
    for event in pygame.event.get():
        if event.type == pygame.QUIT:
            raise SystemExit

        if event.type == pygame.KEYDOWN:
            cur_pressed.add(event.key)

        if event.type == pygame.KEYUP:
            cur_released.add(event.key)
    
    for i in letters:
        if (i in cur_pressed) != pygame.key.get_just_pressed()[i]:
            raise RuntimeError(f"{i}")

        if (i in cur_released) != pygame.key.get_just_released()[i]:
            raise RuntimeError(f"{i}")

    pygame.time.wait(10)

Just run this script and spam letters randomly. There should be no error raised if the implementation is alright.

I'll wait for someone else to merge this PR though, because the last change (as of now) is mine