rac-schema-codegen

name

rac-schema-codegen

description

Update RAC protocol schema TOML under `schemas/rac/*.toml`, map completed RAC commands from `docs/rac/messages` and `artifacts/rac/v16/format_parse_tasklist.md`, and regenerate codegen via `scripts/rac/rac_codegen.py` or `scripts/rac/rac_codegen_all.py`. Use when adding or correcting RPC/response definitions, record layouts, or method IDs; always use inline request fields under `[rpc.*]` (never `[request.*]`), and always set minimal `version` for fields and RPCs.

RAC Schema & Codegen

Overview

Define or update RAC protocol schemas and regenerate the corresponding Rust codegen with the repository tooling.

Workflow

1. Identify completed commands from artifacts/rac/v16/format_parse_tasklist.md and the matching format docs in docs/rac/messages/.
2. Update the relevant schemas/rac/*.toml with sequential record layouts and inline RPC fields (no [request.*] blocks).
3. Regenerate code:
   • Single schema: python3 scripts/rac/rac_codegen.py schemas/rac/<name>.toml --out apps/rac_protocol/src/commands/<name>_generated.rs
   • All schemas: python3 scripts/rac/rac_codegen_all.py
4. If implementation behavior changes, run cargo test -p rac_protocol (expect missing fixture errors if artifacts are absent).

Schema Rules

• Always define request fields inline under [rpc.*] using fields = [...]. Never create [request.*] blocks.
• Responses belong in [response.*]; generate a response struct only when body.struct = true.
• Use field on response bodies to control list/record field names.
• Set minimal version for every field and every RPC entry.
• Never use *_default field types; always specify concrete types (e.g., u32_be, str8).
• Keep schemas codegen-friendly: avoid super:: references and minimize manual types.
• Parsing must be strictly sequential and based on RecordCursor; do not rely on scan helpers.

References

• Read references/schema-format.md for the canonical TOML layout and examples (including the inline RPC/request nuance).