Description

1. ACIR simulator tracks pending notes and allows Noir contracts to "get" them
2. Kernel rejects all transient reads for now (until read<>commitment matching logic is there)
3. New test contract added along with ACIR simulator tests and e2e tests for pending commitments
4. New type ReadRequestMembershipWitness added which is a normal membership witness with a "kernel hint"
   • may need more than a single hint in the future
5. Temporary hack in Noir to force strict dependency and ordering between notifyCreatedNote and subsequent getNotes oracles

Closes tickets...

• Closes [TS] RPC Client (or unconstrained C++ helper) should detect which readRequests read "pending" commitments & flag transient #835
• Closes [with a caveat] [TS] Aztec Noir functions should be able to read pending commitments from new list maintained in ACIR Simulator #919
  • Noir contracts cannot rely on sane ordering of oracle calls until stepwise execution of ACIR opcodes is enforced
    • fix: maintain Opcode ordering in CSATTransformer noir-lang/acvm#419
    • chore: Experimenting with stepwise ACVM solver [DO NOT MERGE] noir-lang/noir#1729

Followup tasks

• [TS / Simulator] Enforce sorting/filtering for pendingNotes #1030
• [TS, Simulator] remove newNotePreimages as it is redundant in light of pendingNoteData #1040
• getCommitment should be able to work with pending commitments #1029
• Aztec Noir functions should be able to nullify pending commitments from list maintained in ACIR Simulator #920
• [Noir, TS, C++] App circuit should accept start_counter, maintain inner counter, and output end_counter #864
• Remove historic private data root from app circuit public inputs - add to PrivateCallData #778

Other Related Tickets

• Relates to Spike: what if app circuit outputs different #read-requests vs what the RPC client gets from the simulator? #779
• Relates to Aztec Noir functions should be able to nullify pending commitments from list maintained in ACIR Simulator #920
• Relates to [Noir, TS, C++] App circuit should output the commitment that a nullifier corresponds to #863
• Relates to Noir contracts shouldn't get notes that were nullified earlier in TX #940

Checklist:

• I have reviewed my diff in github, line by line.
• Every change is related to the PR description.
• I have linked this pull request to the issue(s) that it resolves.
• There are no unexpected formatting changes, superfluous debug logs, or commented-out code.
• The branch has been merged or rebased against the head of its merge target.
• I'm happy for the PR to be merged at the reviewer's next convenience.