feat: implement SetIfPresent

[arthhhhh23]

Added to all implementations and most tests

SetIfPresent sets the provided key to the provided value if it already exists in the cache implementation. For MapTTLCache, it also reinitializes the TTL

[egregors]

Could you please provide some example, why would you want to do that? And as far as you extend public iface, it probably should be represented in readme

[C-Pro]

Looks good to me, thanks.

[arthhhhh23]

Could you please provide some example, why would you want to do that? And as far as you extend public iface, it probably should be represented in readme

Updated the readme!
In terms of examples, one could use it whenever an atomic get-and-then-set is required. Let's say I'm storing a mapping between (user id, user info) where I have 2 processes, one that updates the user info and one that deletes the key if the user logs out. I might do:

userInfo := m.Get(userId)
userInfo.Nickname = updatedNickname
// some other process deletes the userId key
m.SetIfPresent(userId, userInfo)

without SetIfPresent, we could re-insert the user info even though it's supposed to have been deleted

[egregors]

minor: I believe, that developer related stuff (ide settings, personal tooling, etc.) should be ignored in git config --global, not in the project .gitignore. It helps maintain simple, clean, and easy to read and understand ignore file. Project .gitignore could contain some build artifacts, or dot-secrets, or smth like so.

[C-Pro]

Oh, I just realised that it would be a shame to implement atomic Get/Set without returning the previous value. This will increase the number of possible use cases.

[arthhhhh23]

Good idea, I'll do this later

[arthhhhh23]

refactored

[C-Pro]

LGTM, thanks!

[arthhhhh23]

@C-Pro when you have time, could you please merge this and create a new package version? I will likely need this next week