flaky behaviour for the .has method about keyv

[jaredwray]

@elliot40404 - this was a pretty nasty bug and we added in more code to handle all the use cases. Here all all the cases we now handle on has and it will be released in v5 when alpha goes live in the next couple weeks.

• When it is Map() the has does not handle expiring so we need to detect that type and handle it for that case.
• If it is a Keyv Storage Adapter we need to assume they will handle the expiring correctly. Note, we found a bug in @keyv/sqlite and have plans to fix that also in the next 30 days where it doesnt handle the expires.
• If it does not have a has function on the storage adapter then we handle it all ourselves and check the expires

Nice find!

[jaredwray]

@elliot40404 - thanks for sending this over and I will be working to test this and if if so get a fix out with v5 of keyv which should be in alpha this quarter.

[elliot40404]

Hey @jaredwray, thanks for looking into this. I can take this on if you like and create a PR.

[jaredwray]

That would be great. Do you want to work on main branch for this?

[elliot40404]

Yes I can work on main, but if you have any other branch on mind please do let me know.