cln-plugin: add a hook builder to make use of before, after and filters

[daywalker90]

Mirroring 60a0ed2

Changelog-None

• The changelog has been updated in the relevant commit(s) according to the guidelines.
• Tests have been added or modified to reflect the changes.
• Documentation has been reviewed and updated as needed.
• Related issues have been listed and linked, including any that this PR closes.

[cdecker]

ACK 9a6486e

[cdecker]

Is there a point in keeping the HookBuilder around? I quite like using mut HookBuilder as argument type, because it means we own the instance, and we can disect it however we like, so in this case I think we can skip the hook.name.clone() invokation and directly use hook.name, saving us a clone in normal operations 🤔

Very minor thing, but I was wondering what you think of such an optimization.

[daywalker90]

Afaik we don't keep it around since hook.build() consumes it. Since we need the String in name here and to make the Hook in build() we need to clone() it somewhere, right?

[cdecker]

That's the thing that I'm wondering, if we consume the HookBuilder, the String is free to be reassigned, and the Builder could reuse the String. No idea if this is done by the compiler, and more of a curiosity to be honest, so this should absolutely not hold up progress. This can be merged as is.

[daywalker90]

I think about every clone alot aswell 😄 . self.hooks has the hook.name as key and in the value struct. Both are owned so we must either

• clone the name for the key or the value
• use borrowed objects (then we probably need to start using lifetimes, which seems not worth it here imo)
• refactor some stuff so we don't have the name in both key and value

[madelinevibes]

It seems this one missed 25.12 because it wasn't linked to an issue :( added it to the milestone for the next release

[daywalker90]

rebased after #8768

[daywalker90]

CI timed out in pre-build checks in "Check source" at

contrib/pyln-proto/tests/test_invoice.py ..                              [ 13%]
Error: The operation was canceled.

no clue why. It works locally.

[madelinevibes]

@daywalker90... this update re: CI from @ShahanaFarooqui in another forum might explain?

The CI is still failing, but I merged your PR since it had already passed the checks. The workflow will require further debugging, but my best guess is that the GitHub container is running out of resources due to our recently added --reruns=10 pytest option. Please disregard any CI failures unrelated to your changes until the workflow stabilizes.

[ShahanaFarooqui]

Yes, CI is consistently failing the pre-build check for all PRs, even though it passes locally (albeit taking ~100 minutes).