feat: initial implementation of flashblocks sub

[steph-rs]

Proposed solution, addressing #118

Implementation:

• New base_subscribe RPC endpoint with newFlashblocks subscription kind
• Returns full RpcBlock on each flashblock update (similar to eth_subscribe("newHeads"))
• Follows reth's EthPubSub patterns using jsonrpsee subscriptions and broadcast channels
• Only available when flashblocks enabled via --websocket-url flag

[haardikk21]

I wonder if instead of a new RPC method, we should override eth_subscribe with special handling in case params includes the newPendingTransactions subscription type

Edit: nevermind, i guess that doesn't allow consumers to use logs as the subscription type then. we want logs based off of flashblock data

Edit 2: i just saw your comment on the issue

[steph-rs]

@danyalprout please let me know what you think🙏

[haardikk21]

Could you add some tests for this? Probably in tests/rpc.rs

[haardikk21]

cc @cody-wang-cb to take a look at this

[cb-heimdall]

✅ Heimdall Review Status

Requirement	Status	More Info
Reviews	✅ 1/1	Denominator calculation Show calculation 1 if user is bot 0 1 if user is external 0 2 if repo is sensitive 0 From .codeflow.yml 1 Additional review requirements Show calculation Max 0 0 From CODEOWNERS 0 Global minimum 0 Max 1 1 1 if commit is unverified 0 Sum 1

[refcell]

I've ported these changes to the new test-utils setup. Won't approve since I've made a bunch of changes that could use another set of eyes.

[steph-rs]

now we are talking! @refcell it's nice to see new contributors here🔥

[refcell]

now we are talking! @refcell it's nice to see new contributors here🔥

Thx, LMK if I missed something and I'll fix it @steph-rs

[danyalprout]

This looks great to me. Thank you @steph-rs and @refcell!

I personally prefer the idea overriding the existing eth_subscribe if possible. I wonder if we could have eth_subscribe(ExtendedSubscriptionKind), where ExtendedSubscriptionKind = {...SubscriptionKind, NewFlashblocks, PendingLogs }.

This would:

1. Keep it consistent with the rest of the Flashblock API's (that override the underlying eth_ RPC's)
2. Should be backwards compatible with the existing eth_subscribe setup
3. We could potentially override the existing newPendingTransactions event to use the Flashblocks pending state

Happy for us to merge this now and do a follow up to see if we can move it into the eth namespace.

[steph-rs]

I would be happy to try my best in follow-up. should we update the issue then?

[refcell]

I would be happy to try my best in follow-up. should we update the issue then?

How about we just open up a new issue and point the old issue to the new one? And yes, doing this in a follow-on PR will make it easier to prevent more git conflicts - if you could work in a follow-on pr that would be much appreciated!!