feat: refactor append_tx_effects_for_blob

[MirandaWood]

Refactors the append_tx_effects_for_blob function for #11058:

• Removes expensive public data update gathering function
• Combines the two get_effects functions into one to avoid doubling code

Should probably be merged after #11686, since that would be easier to merge into this PR rather than vice versa.

[github-actions]

Changes to circuit sizes

Generated at commit: 6b66a9704202c5396c1ab8f4e1b95a51879c88d7, compared to commit: 438c905e710b54cc194da9be2da561c9a8c27c28

🧾 Summary (100% most significant diffs)

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
rollup_base_private	-2 ✅	-0.00%	-2 ✅	-0.00%
rollup_base_public	-22,155 ✅	-25.73%	-75,073 ✅	-6.42%

---

Full diff report 👇

Program	ACIR opcodes (+/-)	%	Circuit size (+/-)	%
rollup_base_private	137,828 (-2)	-0.00%	1,637,357 (-2)	-0.00%
rollup_base_public	63,937 (-22,155)	-25.73%	1,094,488 (-75,073)	-6.42%

[TomAFrench]

I think this harms readability more than it helps tbh. I would prefer to just remove get_all_update_requests_for_tx_effects and avoid the other changes the but I'm gonna defer to circuits team on this as it's not my codebase.

[MirandaWood]

I think this harms readability more than it helps tbh. I would prefer to just remove get_all_update_requests_for_tx_effects and avoid the other changes the but I'm gonna defer to circuits team on this as it's not my codebase.

Ah ok, I probably misunderstood the linked issue. I assumed 'repeated code' was also referring to the main fn and helper having essentially the same logic, but with either a check or assign. I can remove this, no worries! Will also get a review from the team

[MirandaWood]

I think this harms readability more than it helps tbh. I would prefer to just remove get_all_update_requests_for_tx_effects and avoid the other changes the but I'm gonna defer to circuits team on this as it's not my codebase.

Just FYI have reverted the other changes, lmk on the issue if anything else should be done to this function!