blockifier: add predicted_alias_storage_entries

[yoavGrs]

Predicts the alias-contract storage entries that allocate_aliases_in_storage
would touch for a given state_diff, without running the allocation. Used by
the batcher to assemble the accessed-keys set written for OS replay.

Co-Authored-By: Claude Opus 4.7 (1M context) noreply@anthropic.com

[reviewable-StarkWare]

This change is 

[yoavGrs]

• starknet_os_flow_tests: drive commitment keys from compute_accessed_keys #14106
• apollo_storage,apollo_batcher,apollo_reverts: write accessed keys instead of tx execution infos #14105
• blockifier: take CommitmentStateDiff in AccessedKeys::new and predicted_alias_storage_entries #14156
• apollo_batcher: record ProofFacts block_number per tx #14104
• blockifier: add compute_accessed_keys and AccessedKeys #14103
• blockifier: add predicted_alias_storage_entries #14102 👈 (View in Graphite)
• blockifier: add CallInfo::get_visited_contract_addresses #14101
• apollo_integration_tests: add shared proof-flow infrastructure (#13760) #14100
• main

This stack of pull requests is managed by Graphite. Learn more about stacking.

[cursor]

PR Summary

Low Risk
Low risk: adds a pure helper for predicting alias-contract storage access plus a parity test; core alias allocation/compression behavior is unchanged aside from refactoring the compression predicates.

Overview
Adds predicted_alias_storage_entries, a helper that computes which alias-contract (ContractAddress, StorageKey) entries allocate_aliases_in_storage will touch for a given StateMaps, without performing allocation.

Refactors the compression decision logic into shared should_compress_address/should_compress_storage_key helpers, and adds a unit test asserting the predicted entry set matches actual alias-contract writes when no aliases pre-exist.

^{Reviewed by Cursor Bugbot for commit b5f856e. Bugbot is set up for automated code reviews on this repo. Configure here.}

[ArielElp]

@ArielElp reviewed 2 files and all commit messages, and made 1 comment.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on yoavGrs).

[dorimedini-starkware]

@dorimedini-starkware reviewed 2 files and all commit messages, and made 1 comment.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on yoavGrs).

---

crates/blockifier/src/state/stateful_compression.rs line 85 at r1 (raw file):

/// read (and may write) for the given `state_diff`, without running the allocation. Mirrors the
/// access pattern of `AliasUpdater::new` (counter read) plus each `AliasUpdater::insert_alias`
/// call (one read per qualifying key).

kind of annoying that you need this function. can you see if you can reuse / generalize allocate_aliases_in_storage for this?
in your test it seems that they seem to have the same result...

I would go on to say the allocate_aliases_in_storage should be an associated function of AliasUpdater, and the new predicted_alias_storage_entries should also be an associated function of AliasUpdater.

(it bugs me because there are duplicated ifs and scanning logic, I prefer these to be in one place..)

Code quote:

/// Predicts the storage entries on the alias contract that `allocate_aliases_in_storage` will
/// read (and may write) for the given `state_diff`, without running the allocation. Mirrors the
/// access pattern of `AliasUpdater::new` (counter read) plus each `AliasUpdater::insert_alias`
/// call (one read per qualifying key).

[yoavGrs]

@yoavGrs made 1 comment.
Reviewable status: all files reviewed, 1 unresolved discussion (waiting on dorimedini-starkware).

---

crates/blockifier/src/state/stateful_compression.rs line 85 at r1 (raw file):

Previously, dorimedini-starkware wrote…

kind of annoying that you need this function. can you see if you can reuse / generalize allocate_aliases_in_storage for this?
in your test it seems that they seem to have the same result...

I would go on to say the allocate_aliases_in_storage should be an associated function of AliasUpdater, and the new predicted_alias_storage_entries should also be an associated function of AliasUpdater.

(it bugs me because there are duplicated ifs and scanning logic, I prefer these to be in one place..)

Done.

[github-actions]

Artifacts upload workflows:

• Started at 2026-05-20T13:46:24Z
• Started at 2026-05-20T13:46:19Z

[dorimedini-starkware]

@dorimedini-starkware reviewed 1 file and all commit messages, and resolved 1 discussion.
Reviewable status: complete! all files reviewed, all discussions resolved (waiting on yoavGrs).