Define the generated corpus file schema and layout #39

New issue

Closed

opened 2026-07-01 00:51:45 +00:00 by manlycucumber · 0 comments

manlycucumber commented

2026-07-01 00:51:45 +00:00

(Migrated from github.com)

Problem

The engine reads generated static data. A documented schema and on-disk layout for identity records, per-edition attributes, temporal definitions, and precedence data is needed before generation can begin.

Design alignment (v0.1.0)

The corpus follows the identity / edition / year model. The identity record carries only the Observance slug (<rite>:<cycle>:<body>), kind, titulars, names, and aliases — no rank, colour, octave, or date. Rank / colour / placement live in per-edition attribute overlays keyed by (id, edition). Temporal day IDs are structural only (no Easter-offset aliases). See docs/design/observance-id-grammar.md.

Acceptance criteria

Schema documents the identity record and the per-edition attribute overlay separately
Identity record contains no rank / colour / date (those are edition attributes)
On-disk layout (paths, file format) is specified
A JSON schema or equivalent validates a sample file
Each text string carries a CC0 provenance marker

Notes

Corpus is CC0; text strings carry per-string provenance (ties to the SOURCES ledger).

## Problem The engine reads generated static data. A documented schema and on-disk layout for identity records, per-edition attributes, temporal definitions, and precedence data is needed before generation can begin. ## Design alignment (v0.1.0) The corpus follows the **identity / edition / year** model. The **identity record carries only** the `Observance` slug (`<rite>:<cycle>:<body>`), kind, titulars, names, and aliases — **no rank, colour, octave, or date**. Rank / colour / placement live in **per-edition attribute** overlays keyed by (id, edition). Temporal day IDs are **structural only** (no Easter-offset aliases). See `docs/design/observance-id-grammar.md`. ## Acceptance criteria - [ ] Schema documents the identity record and the per-edition attribute overlay **separately** - [ ] Identity record contains **no** rank / colour / date (those are edition attributes) - [ ] On-disk layout (paths, file format) is specified - [ ] A JSON schema or equivalent validates a sample file - [ ] Each text string carries a CC0 provenance marker ## Notes Corpus is CC0; text strings carry per-string provenance (ties to the SOURCES ledger).