perf: drop a ~67 KiB duplicate prost decode tree + hoist a per-message traversal

[jlucaso1]

Two small, no-drawback perf cleanups found by auditing for binary bloat and per-message waste.

1. Binary: reuse the canonical buffer-type for MessageContextInfo merge (~67 KiB .text)

MessageUtils::encode_dm_plaintexts folds reporting context into a message's MessageContextInfo via ctx.merge(vec.as_slice()). The bare &[u8] buffer made prost monomorphize the entire MessageContextInfo → BotMetadata decode subtree in a distinct buffer-type shape, separate from the &mut &mut &[u8] tree that Message::decode already instantiates — so the binary shipped a second copy of that subtree (BotMetadata::merge_field alone is ~51–67 KiB).

Fix: route through a pinned codec::message_context_info_merge that merges via &mut &mut &[u8], matching the shape the rest of the workspace decodes with (the waproto::codec module already exists for exactly this "pin one instantiation" purpose).

Verified locally with nm on the release artifact — each nested type drops from two merge_field copies to one:

symbol	before	after
BotMetadata::merge_field	2 copies (&mut &mut &[u8] + &[u8])	1 (&mut &mut &[u8])
MessageContextInfo::merge_field	2	1
ContextInfo::merge_field	2	1

Same merge semantics (later-set fields win); this is a cold path (only the rare case where the message already carries a top-level message_context_info). No behavior change.

2. Runtime: hoist a duplicated traversal in dispatch_parsed_message

On every inbound message, msg.get_base_message().get_ephemeral_expiration() was computed twice — once in the if condition, once in the assignment. Each call walks the wrapper levels and iterates the ~28-variant content list. Bind it once with a let-chain. Tiny, but pure waste on the hottest path.

---

Both ran cargo fmt/clippy clean and the wacore message tests pass locally. Leaving the full suite + the binary-size CI report (which should show a .text decrease) + CodSpeed to CI.

Context: this came out of a fresh audit after the appstate dedup PR (#868); the remaining heavy benchmarks are genuine crypto/zlib, so these target binary bloat and per-message overhead rather than a hot algorithmic loop.

---

Generated by Claude Code

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: b9099f63-96e4-49c5-9671-35687f0a0a7b

📥 Commits

Reviewing files that changed from the base of the PR and between 9cfa4db and 4a0c4e9.

📒 Files selected for processing (3)

• src/message/dispatch.rs
• wacore/src/messages.rs
• waproto/src/lib.rs

---

📝 Walkthrough

Summary by CodeRabbit

• Refactor
  • Internal improvements to message expiration handling with simplified logic.
  • Enhanced message context information processing for better efficiency.
  • Optimized code organization and test-specific configurations.

Walkthrough

Adds message_context_info_merge to waproto::codec for merging protobuf bytes into an existing MessageContextInfo. wacore/src/messages.rs adopts this helper in encode_dm_plaintexts and moves the prost::Message import behind #[cfg(test)]. src/message/dispatch.rs simplifies ephemeral_expiration extraction to a single pattern match.

Changes

MessageContextInfo Merge Helper and Dispatch Cleanup

Layer / File(s)	Summary
New message_context_info_merge codec helper and adoption waproto/src/lib.rs, wacore/src/messages.rs	waproto::codec gains pub fn message_context_info_merge that merges protobuf bytes into an existing MessageContextInfo via prost. encode_dm_plaintexts switches from ctx.merge(...) to this new helper. The prost::Message import is scoped behind #[cfg(test)].
Ephemeral expiration pattern-match simplification src/message/dispatch.rs	dispatch_parsed_message replaces the is_some() guard plus second get_ephemeral_expiration() call with a single let Some(exp) destructure, assigning via Arc::make_mut.

Estimated code review effort

🎯 2 (Simple) | ⏱️ ~8 minutes

Possibly related PRs

• oxidezap/whatsapp-rust#520: Directly overlaps with the Arc<MessageInfo>/Arc::make_mut-based ephemeral mutation refactor in dispatch_parsed_message — same function, same field.
• oxidezap/whatsapp-rust#787: Both PRs touch MessageUtils::encode_dm_plaintexts in wacore/src/messages.rs around the message_context_info merge logic.
• oxidezap/whatsapp-rust#842: The waproto::codec::message_context_info_merge helper introduced here is directly tied to the codec work from that PR.

Suggested labels

performance

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly and accurately summarizes both main performance improvements: binary bloat reduction via prost decode tree deduplication and the per-message ephemeral expiration traversal optimization.
Description check	✅ Passed	The description is comprehensive and directly related to the changeset, providing technical justification, verification details, and context for both performance improvements.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch claude/whatsapp-rust-pr-review-mz0gyy

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[cubic-dev-ai]

No issues found across 3 files

<sub>Tip: cubic could auto-approve low-risk PRs like this, if it thinks it's safe to merge. Learn more

Re-trigger cubic</sub>

[github-actions]

📦 Binary size report

Metric	main	PR	Δ
bin size (stripped)	10.67 MiB	10.58 MiB	-97.84 KiB (-0.90%) 🔽
bin .text	8.76 MiB	8.67 MiB	-95.94 KiB (-1.07%) 🎉
bin allocated (text+data+bss)	10.67 MiB	10.58 MiB	-96.00 KiB (-0.88%) 🔽
llvm-lines wacore	666,701	644,384	-22,317 (-3.35%) 🎉
llvm-lines wacore copies	18,047	17,673	-374 (-2.07%) 🎉
llvm-lines whatsapp-rust lib	652,742	652,742	0
llvm-lines whatsapp-rust lib copies	19,857	19,857	0
deps crates (Cargo.lock)	354	354	0

.text per crate

Crate	main	PR	Δ
.text whatsapp_rust	1.47 MiB	1.47 MiB	-760 B (-0.05%) 🔽
.text wacore	556.04 KiB	543.77 KiB	-12.28 KiB (-2.21%) 🎉
.text wacore_binary	103.64 KiB	103.64 KiB	0
.text wacore_libsignal	168.32 KiB	168.32 KiB	0
.text wacore_appstate	35.26 KiB	35.26 KiB	0
.text wacore_noise	30.68 KiB	30.68 KiB	0
.text waproto	960.72 KiB	895.34 KiB	-65.38 KiB (-6.81%) 🎉
.text whatsapp_rust_sqlite_storage	206.21 KiB	206.21 KiB	0
.text whatsapp_rust_tokio_transport	33.09 KiB	33.09 KiB	0
.text whatsapp_rust_ureq_http_client	6.19 KiB	6.19 KiB	0
.text std	1.13 MiB	1.13 MiB	+72 B (+0.01%) 🔺
.text other deps	4.04 MiB	4.02 MiB	-17.36 KiB (-0.42%) 🔽

Top movers (cargo-bloat attribution)

Crate	main	PR	Δ
waproto	960.72 KiB	895.34 KiB	-65.38 KiB (-6.81%)
prost	480.43 KiB	464.26 KiB	-16.17 KiB (-3.37%)
wacore	556.04 KiB	543.77 KiB	-12.28 KiB (-2.21%)
rustix	131 B	1.41 KiB	+1.28 KiB (+1002.29%)
regex_automata	4.16 KiB	2.88 KiB	-1.28 KiB (-30.81%)
bytes	7.74 KiB	6.55 KiB	-1.19 KiB (-15.33%)

Baseline: 9cfa4dba3 (latest main run) · Head: 820c04e64 · Graphs

[codspeed-hq]

Merging this PR will not alter performance

✅ 172 untouched benchmarks

---

_{Comparing claude/whatsapp-rust-pr-review-mz0gyy (4a0c4e9) with main (9cfa4db)}