[AdaWorldAPI]

The seam this fills was already cut

ogar-loco's own module doc names it: the crate is "everything that is the same no matter what the bytes mean", with "a vocabulary per domain, declared by whoever owns it" plugging into the Vocabulary trait. R2IL is such a domain. This crate mints no node layout, lane carving, call encoding or second addressing system — it is the sibling codebook that seam exists for.

Always on: no feature gate, no optional = true, no cfg. A vocabulary that must be switched on is one a caller can forget to switch on. R2ILVocabulary is zero-sized, so the cheapness is real rather than asserted.

No r2sleigh dependency, deliberately

What loco needs from r2il's opcode enum is an arity table. Depending on the crate would drag its Varnode/SpaceId/ArchSpec object graph into every consumer that only wanted to know that IntAdd pops two — the graph ruff_r2il's own doc calls "structurally still stage-1 pointer chasing". The coupling is by name, guarded by a test that parses r2il/src/opcode.rs.

The table is derived, and the first draft proves why that matters

I wrote it from memory first. Against an enum of 82 it had 77 entries — five variants missing (LoadGuarded, Subpiece, Float2Int, Multiequal, Extract, …) and nine invented (Switch, Halt, AtomicRmw, …). That is the invented-data failure the data-as-config rule exists for, caught before it shipped.

The shipped table is generated from the enum's own field structure: arity = count of Varnode input fields (dst/output/result excluded, being the push side); pushes_result = presence of one.

None means variadic, not zero. CallOther and Multiequal carry inputs: Vec<Varnode> — their operand count is a per-site property no table can hold. None is Vocabulary's own "not covered (refused)", so a body using them is refused rather than silently treated as nullary, which a 0 would have meant.

Arithmetic is not aliased onto the shared core's ADD. The core's ADD is loco's semantics; R2IL's IntAdd carries a machine's flag/width behaviour the core never promised. Same word, two contracts — aliasing them is the "same name, different semantics" trap I-LEGACY-API-FEATURE-GATED catalogues five instances of.

The masked lane projection

call_in_slab already reads call i under a shape — the reshuffle primitive existed. This adds the masked, lazy form:

project(&slab, shape, &mask) -> impl Iterator<Item = (u32, Call)>

Nothing outside the mask is decoded; no Vec<Call> is built. CallMask is Box<[u64]> over call indices with the same algebra and the same single named materializer as every other mask plane in this workspace. Indices are shape-relative, so the shape is carried rather than inferred.

Reshuffling is a re-READ, never a re-WRITE — project borrows &[u8] and returns an iterator, so the same slab under two shapes is two reads of unchanged bytes. A shape is a lens, not a migration, and the_same_bytes_project_differently_under_two_shapes pins exactly that.

Falsifiers

10 tests. Five disable runs, each verified red-then-green:

falsifier	disable
the_table_matches_r2il_s_own_enum	one mnemonic reverted to the invented name
variadic_opcodes_are_refused_rather_than_called_nullary	None → unwrap_or(0)
a_complement_never_invents_call_slots	tail-word masking removed
the_projection_yields_exactly_the_masked_calls	mask ignored, project yields all
a_body_with_no_r2il_calls_masks_to_nothing	vocabulary claims shared-core bytes

The drift test was confirmed to actually parse r2sleigh rather than hit its skip-if-absent path, and asserts a plausible variant count (>40) before comparing — so a moved file degrades to a loud skip instead of a silent pass.

Gates

cargo test -p ogar-r2il 10/10 · clippy --all-targets -D warnings clean · fmt --check clean.

🤖 Generated with Claude Code

https://claude.ai/code/session_01DCfrD5y19cvFc4AoyydXYv

---

Generated by Claude Code

[coderabbitai]

Important

• 🔍 Trigger review

This repository does not receive automatic reviews because it has fewer than 10 stars.

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: da0a2e8b-3cee-4199-a4b4-d7270afa8768

---

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.}

[cursor]

Bugbot couldn't run - usage limit reached

Bugbot is counted against Cursor usage for this user or team, and this run hit a usage or spend limit.

A user or team admin can review and increase usage limits in the Cursor dashboard.

(requestId: serverGenReqId_60936d32-8a03-4152-b508-d9da37e2d43e)