sqlite async API

[ronag]

Provide an async API to our sqlite bindings. In order to avoid blocking the event loop on lookups that require IO.

[benjamingr]

@ronag can you add "why" to the issue? (You can copy it from the original discussion).

[geeksilva97]

Hey, @ronag . To make this happen, Would there be a DatabaseAsync - in contrast to the DatabaseSync - class?

[geeksilva97]

Hey, @ronag . To make this happen, Would there be a DatabaseAsync - in contrast to the DatabaseSync - class?

This doesn't make so much sense. I think it would be better to have Database just like fs that has someFunction and someFunctionSync.

[cjihrig]

Just FYI, this appears to be the stance of better-sqlite3 on the topic - https://github.com/WiseLibs/better-sqlite3/blob/HEAD/docs/threads.md.

[BurningEnlightenment]

A detailed description of a use case which would benefit from offloading to a thread pool can be found here:

• sqlite: better-sqlite3 shouldn't be the default driver (as is), nor be used if more than one sqlite instance or other DBs are also used kysely-org/kysely#1385

[geeksilva97]

A detailed description of a use case which would benefit from offloading to a thread pool can be found here:

• sqlite: better-sqlite3 shouldn't be the default driver (as is), nor be used if more than one sqlite instance or other DBs are also used kysely-org/kysely#1385

IMHO, this link clarifies the use cases and the necessity for an async API.

That said, I'd love to start working on this. What do you say, @cjihrig?

[geeksilva97]

Heads up, I started working on it. I will open PR by next month to get some feedback.

[himself65]

Heads up, I started working on it. I will open PR by next month to get some feedback.

this is awesome. cant wait for that. let me know if you have the PR. i will defenitely test it