Cap ContextColumn.cells at u8::MAX, matching context_columns

[thedavidmeister]

Closes #198

The defect

Expression.context_columns carries #[validate(length(max = "u8::MAX"))]. Its
sibling axis, ContextColumn.cells, carried no length bound at all — one axis of
the same context matrix rejected a 256th entry while the other accepted 100000.

The fix

cells gets the identical length(max = "u8::MAX"). Both axes of the matrix are
now bounded the same way, and the model no longer says two different things about
how big a context matrix may be.

The columns cap is not raised. The issue's other branch — bump 255 to 256
because a u8 index addresses 0..=255 — reads the attribute as a bound on the
index; it is a bound on the count, which is what length(max) measures. Caps in
this org are floor-only: a cap is not raised to admit one more entry. With the two
axes symmetric, the code now shows 255 as the rule rather than as an off-by-one in
one field, which is the "document 255" branch the issue offers.

The only documents this newly rejects are those with more than 255 cells in a
single column. cells appears nowhere in this repo outside this type and its
tests, so nothing here indexes past 255 today.

Relation to the other open issues on this file

• InterpreterCallerMeta: context_columns cap length(max = u8::MAX) rejects 256 columns — off-by-one if the intent is u8-addressable column indices #221 is this same cap from the columns side, asking whether 255 should become
  256. This PR answers no, and leaves the context_columns attribute and both of
  its boundary tests (test_length_constraints, test_context_columns_max_255)
  untouched — so InterpreterCallerMeta: context_columns cap length(max = u8::MAX) rejects 256 columns — off-by-one if the intent is u8-addressable column indices #221 needs no code change, only the ruling above.
• Method.expressions permits an empty array despite 'at least one expression' docs #167 (Method.expressions permits an empty array) and InterpreterCallerMeta.desc doc comment is a copy-paste of abi_name's ("Name of the caller..."), plus 'feild' typos #196 (desc doc
  comment is a copy-paste of abi_name's, plus the "feild" typos) touch the same
  file but different lines. Neither is in this diff and neither conflicts with it.

QA

• Discriminating tests: meta::types::interpreter_caller::v1::tests::test_length_constraints, extended with the cells boundary (255 cells in one column validate, 256 fail) — fails on base, verified by deleting the new attribute at v1.rs:145 and rerunning: called Result::unwrap_err() on an Ok value: InterpreterCallerMeta { ... } panicking at v1.rs:348, the 256-cell assertion. Attribute restored, re-run green (17 passed, 0 failed).
• Mutations applied: v1.rs:145 remove #[validate(length(max = "u8::MAX"))] from ContextColumn.cells (= the pre-fix body) -> killed by test_length_constraints. v1.rs:127 the same removal on Expression.context_columns is outside this diff and already killed by test_length_constraints and test_context_columns_max_255.
• Oracle: the type's own declared bound, not the implementation — context_columns states the context-matrix axis bound as "the count fits in a u8" (255), and cells is the other axis of that same matrix, so its expected boundary is 255 pass / 256 fail regardless of what the code did.
• Category check: the issue asks two things — (1) the 255-vs-256 columns cap, (2) the missing cells cap. (2) is fixed here with the boundary test above. (1) is ruled rather than changed: it is a count cap and caps are not raised, so context_columns and its two boundary tests are deliberately untouched and the symmetry added here is the issue's own "document 255" branch. Both covered.
• Not run locally: the full suite (shared machine) — CI covers it. Run locally: cargo test -p rain-metadata --lib meta::types::interpreter_caller::v1 (17 passed), cargo fmt --check -p rain-metadata, cargo clippy -p rain-metadata --lib --all-targets -- -D warnings (clean).

[coderabbitai]

Warning

Review limit reached

Next included review available in 12 minutes.

View limit details

Limit details: You’ve used the included review currently available.

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

Learn how review limits work.

Review configuration:

⚙️ Run configuration

Configuration used: Organization UI

Review profile: ASSERTIVE

Plan: Pro Plus

Run ID: ab455f9d-3a75-485d-9be6-d8f0b2e3f297

📥 Commits

Reviewing files that changed from the base of the PR and between 45ca96c and f67d202.

📒 Files selected for processing (1)

• crates/cli/src/meta/types/interpreter_caller/v1.rs

---

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.}