Ingest the Clementine Vulgate psalter and canticles #112

New issue

Closed

opened 2026-07-01 00:54:24 +00:00 by manlycucumber · 1 comment

manlycucumber commented

2026-07-01 00:54:24 +00:00

(Migrated from github.com)

Problem

The Latin office draws psalms and canticles from the Clementine Vulgate. These must be ingested, keyed by psalm/verse, with provenance, for assembly.

Acceptance criteria

Clementine Vulgate psalms and canticles ingested and keyed
Provenance and CC0/public-domain status recorded
Verse-level keys align with office references
Spot-check confirms a sample psalm text

Notes

Public-domain Latin source.

## Problem The Latin office draws psalms and canticles from the Clementine Vulgate. These must be ingested, keyed by psalm/verse, with provenance, for assembly. ## Acceptance criteria - [ ] Clementine Vulgate psalms and canticles ingested and keyed - [ ] Provenance and CC0/public-domain status recorded - [ ] Verse-level keys align with office references - [ ] Spot-check confirms a sample psalm text ## Notes Public-domain Latin source.

manlycucumber commented

2026-08-05 19:53:02 +00:00

Owner

Done. The Clementine psalter and the five Office canticles are ingested as source-scoped, verse-keyed streams:

04f37a2 — ingestion core: NFC normalization, verse-keying, fail-closed dedup (tools/generator/src/ingest.mjs)
bebb2a4 — Clementine + Douay-Rheims psalters → data/corpus/text/clementine-vulgate/psalms.ndjson
51a683a — the five Office canticles → canticles.ndjson

ACs: psalms/canticles ingested and keyed ✓; PD status recorded — clementina-1592 is registered in the enforced tools/generator/facts/sources.yaml and carries 2,603 citations ✓; verse keys align with office references via the edition-invariant psalm:verse key ✓; spot-check — every verse was re-derived token-for-token against the raw sha-pinned source, and tools/generator/test/parse-psalter.test.mjs covers the parse ✓.

Done. The Clementine psalter and the five Office canticles are ingested as source-scoped, verse-keyed streams: - `04f37a2` — ingestion core: NFC normalization, verse-keying, fail-closed dedup (`tools/generator/src/ingest.mjs`) - `bebb2a4` — Clementine + Douay-Rheims psalters → `data/corpus/text/clementine-vulgate/psalms.ndjson` - `51a683a` — the five Office canticles → `canticles.ndjson` ACs: psalms/canticles ingested and keyed ✓; PD status recorded — `clementina-1592` is registered in the enforced `tools/generator/facts/sources.yaml` and carries 2,603 citations ✓; verse keys align with office references via the edition-invariant `psalm:verse` key ✓; spot-check — every verse was re-derived token-for-token against the raw sha-pinned source, and `tools/generator/test/parse-psalter.test.mjs` covers the parse ✓.

manlycucumber closed this issue

2026-08-05 19:53:02 +00:00

manlycucumber referenced this issue

2026-08-05 19:53:16 +00:00

[Epic] Latin psalter & scripture sources #111