Documents the full set of changes from this session: the
nocturnReadingIds precedence bug fix, the new Matins hymn Common-tier
and invitatory-antiphon mechanism (Common-of-an-Apostle content
authored, other Commons/seasons logged as backlog), the antiphon-
doubling dedup, and the Terce/Sext/None hymn dedup — including why the
Lauds/Vespers office-bundle resolvers were deliberately left alone
(they turned out not to be byte-identical: different override-
eligibility semantics) and why Matins's own octave hymn tier isn't
redundant with resolveOfficeWinner's octave-folding.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01F25189JqjXddUhU9hM9nUS
Terce, Sext, and None each had a byte-identical 'hymn' case
(resolveCommon(part.textRef.id), no doxology mechanism, no override/
season tiering). Replace all three with a single shared
resolveSelfContainedHymn helper in resolve-common.ts.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01F25189JqjXddUhU9hM9nUS
resolveMatinsHymn gained a Common-category tier (override -> octave ->
Common category -> season -> ferial), reusing matins-psalmody-
overrides.ts's own getSaintRecord(id)?.common lookup pattern. The
invitatory antiphon (invitatoryParts), previously one hardcoded fixed
text for every day of the year with only full-vs-incipit doubling
varying by rank, now goes through a real resolveMatinsInvitatoryText
(override -> Common category -> season -> ferial, no octave tier) before
the doubling step.
Content authored this pass: Common-of-an-Apostle only, both hymn
(matins-hymn-common-of-an-apostle.yml, "Aeterna Christi munera") and
invitatory antiphon (matins-invitatory-common-of-an-apostle.yml, "Regem
Apostolorum Dominum"), both read directly from the reference engine's
Commune/C1.txt. Live-verified: St. Bartholomew (2026-08-24) and St.
Andrew (2026-11-30) both now resolve to this Common's text.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01F25189JqjXddUhU9hM9nUS
The same idiom (splitNamedAntiphon + isDoubleOrHigher(...) ? full :
incipit) was copy-pasted at 11 call sites across 8 hour files. Add one
shared openingAntiphon(antiphon, winner) helper in resolve-common.ts
and use it everywhere a psalm/canticle antiphon's opening form is
decided. Pure dedup, no behavior change.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01F25189JqjXddUhU9hM9nUS
nocturnReadingIds unconditionally pooled the plain temporal-cycle and
month-week reading ids into every day's nocturn-reading pool, even when
a real feast wins the day outright with zero commemorations
(decideOccurrence's `ordinary-feria` branch, which correctly suppresses
the temporal identity entirely). St. Bartholomew (duplex-2-classis,
2026-08-24, a Monday) has no Nocturn 3 content of his own, so the
leftover 13th-Sunday-after-Pentecost and month-week readings wrongly
filled his Nocturn 3.
Gate the pooling on whether the day's own occurrence decision actually
retained the temporal identity (day.winner.kind === 'temporal', or a
`temporal`-kind entry in day.commemorations) — mirrors getDayCollects's
existing pattern for the same question.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01F25189JqjXddUhU9hM9nUS
Matins now sings "Te decet laus" (Psalterium/Common/Prayers.txt's [Te
decet]) immediately after the Te Deum on every 3-nocturn day, matching
Rule of Benedict ch. 11's close of Vigils. New common text
te-decet-laus.yml, wired into both matins.ts branches that render a Te
Deum; absent on ferial days along with it.
Also drops Psalm 129 (De profundis) from the Lauds ferial Preces and
Psalm 50 (Miserere) from the Vespers ferial Preces, per direct
instruction — both were embedded inline in the live-transcribed text.
Updated the affected fixture tests to key off a versicle that survives
the removal instead of the now-absent psalm text.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01EWzE3kaSZCs79fPBrrE1aF
All 53 nocturn-readings files with a responsory field started straight
into the text with no leading R., unlike octave-readings' own R./V.
convention. Fixed in content per direct instruction (not injected at
render time), matching each file's own existing plain V./R. style rather
than switching to octave-readings' R./V. glyphs.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01EWzE3kaSZCs79fPBrrE1aF
buildReadingPool concatenated readings in id-processing order, so a later
post-Pentecost Sunday's Nocturn 3 (from its own temporalId file) could land
ahead of its Nocturn 2 (from the newly-added month-week id) in the pool,
flipping which nocturn slot each landed in. Group by each reading's own
nocturn tag instead, ascending, preserving id/file order within each group.
Also logs three follow-up items to TODO.md: displaying "Te decet laus",
the missing responsory marker convention in nocturn-readings content, and
removing Ps 129/50 from the Lauds/Vespers ferial Preces.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01EWzE3kaSZCs79fPBrrE1aF
The later post-Pentecost Sundays' real Nocturn 2 patristic content isn't
keyed by Pentecost offset in the reference engine — it's keyed by civil
calendar month/week (the old "Scriptura occurrens" cycle), which shifts
relative to post-pentecost-NN every year depending on Easter's date. Port
Date.pm's monthday() algorithm (pre-1955/Tridentine rubrics) as a new,
date-driven id, author all 20 August-November month-week files from the
reference engine's Tempora/0MN-0.txt sources, and pool them in
hours/matins.ts as a third, independent reading source alongside the
existing temporalId lookup.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01EWzE3kaSZCs79fPBrrE1aF
Closes the gap the user found on 2026-08-23 (post-pentecost-13's Matins had
no patristic reading): nocturn-readings/ had content for only one temporal
id (post-pentecost-15, the original mechanism proof date). Since
resolveTemporalId maps every day in a liturgical week to the same id, one
file per id covers the whole week — authored all 23 remaining
post-pentecost-NN ids (01-14, 16-24) from the reference engine's own
Tempora/PentNN-0.txt files. Trinity Sunday (01) keeps its own full proper
office including a displaced-Sunday commemoration; 02-11 have real Nocturn
2+3 content; 12 onward are Nocturn-3-only (homily, Gospel incipit skipped
per convention), reusing the same two responsories the source itself
cross-references week to week.
Rest of the temporal cycle (Advent/Christmas/Epiphany/Septuagesima/Lent/
Easter) is a separate follow-up pass, tracked in TODO.md.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01EWzE3kaSZCs79fPBrrE1aF
Every psalm's opening antiphon (Lauds, Prime, Terce, Sext, None,
Vespers, Compline's callers, Matins) now runs through applyFlexaMark,
marking the boundary in both the displayed antiphon and the psalm's
first verse — e.g. today's Sunday Matins Nocturn 1: "Ant. The king
rejoices ‡" / "The king rejoices ‡ in thy strength, O Lord...".
Live-verified against the reference engine's own rendering for Ps 20.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UCvvgFLrhXNVFUThz6kanw
Ports Divinum Officium's getantcross/depunct (horas.pl): finds where a
displayed antiphon's words stop matching a psalm's first verse, so the
mark can later show where the antiphon leaves off and the psalm's own
text resumes. Not yet wired into any hour builder.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01UCvvgFLrhXNVFUThz6kanw
Researched whether September/Advent Ember days belong in the
movable-feasts.ts table alongside Christ the King/Immaculate Heart of
Mary. Live-verified against the reference engine they don't: Ember days
are privileged ferias (a saint can still win outright against them), not
named feasts contesting the day by rank.
Two smaller pieces instead: calendar/temporal.ts's septemberEmberDayOffset
(Sunday nearest Sept 14, same "nearest" arithmetic adventStart already
used, factored into a shared nearestSunday helper) feeds a new
resolveTemporalCategory check giving those 3 dates privileged-feria-minor
(live-verified correct); calendar/ember-days.ts's applyEmberDay relabels
the day's own temporal id to the Ember day's own (so its real content is
found) when the feria itself wins, or adds a commemoration when a saint
does. Advent Ember days needed no precedence change -- Advent's own
season default already covers them.
Found and fixed a real bug along the way: matins.ts's nocturnReadingIds
only ever pulled a commemorated *sanctoral* id's own readings, never a
commemorated *temporal* one's -- so a commemorated Ember day (the common
case) would never have surfaced its own content. Same root cause as the
day.winner.id gap the IHM relocation fixed earlier, just hiding in the
commemorations loop instead.
Authored all 6 days' real collect + 3 Matins readings each, transcribed
directly from the reference engine (September: Tempora/093-{3,5,6}.txt;
Advent: Tempora/Adv3-{3,5,6}.txt, whose lessons are themselves a
cross-reference to the Annunciation's own Common, followed and
transcribed from there). Kept as 3 separate readings per day rather than
the usual combine-into-one default, since each carries its own
genuinely distinct proper responsory.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VZSAgRi4QE4XRTqVto93zA
Extends TemporalFeastRecord's date anchor from a single easterOffset
number into a MovableAnchor union: 'easter' (unchanged), plus
'nth-sunday-of-month' (nth: -1 for "last") and 'nth-sunday-of-advent' --
both new anchor kinds resolve to "Nth Sunday counting from an anchor
date," just with a different anchor (first-of-month vs. Advent's own
start, which is itself always a Sunday). Adds calendar/temporal.ts's
sundayOnOrAfter to support the month case.
Migrates christ-the-king.yml onto the new nth-sunday-of-month anchor
(month: 10, nth: -1) with a new `unconditional: true` flag replacing its
old always-wins-no-rank-check bespoke function -- retiring
applyChristTheKing and its lastSundayOfOctober helper from
calendar/index.ts entirely. Down to two bespoke override functions left
(marian-saturday's date-less fallback, christmas-octave-sunday's
fixed-date-range Sunday search), both structurally different from the
Nth-Sunday-from-anchor shape this table now covers.
No behavior change -- the full existing Christ the King test suite
passes unchanged through the new generic path.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VZSAgRi4QE4XRTqVto93zA
The bespoke applyImmaculateHeart function (following the applyMarianSaturday/
applyChristTheKing precedent of one hand-written override function per
feast) doesn't scale: several more feasts are planned that all reduce to
"weekday N days from Easter Sunday" (Ember/Rogation days, more
Sacred-Heart-family Marian devotions, St. Joseph's pre-1955 Eastertide
feast, Lenten Friday Passion devotions), and one-off functions invite
subtle ordering bugs -- applyImmaculateHeart had to be sequenced after
applyMarianSaturday specifically or it would have been silently clobbered.
Replaced with calendar/movable-feasts.ts's generic applyMovableFeasts,
which scans calendar/temporal-feasts.ts's existing per-feast YAML records
for a new optional `easterOffset` field. Adding the next movable feast is
now a new data/calendar/temporal-feasts/<id>.yml file with `rank` and
`easterOffset` set, not a new TypeScript function. Same rank-compared,
commemorate-the-loser semantics as before, just centralized instead of
duplicated per feast.
christ-the-king, marian-saturday, and christmas-octave-sunday stay as
their own functions -- they're structurally different rules (a fixed
month-position search, a date-less fallback default, and a fixed-date-
range Sunday search, respectively), not Easter offsets, so force-fitting
them into this table wouldn't actually simplify anything.
No behavior change -- the existing Immaculate Heart of Mary test suite
passes unchanged through the new generic path.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VZSAgRi4QE4XRTqVto93zA
resolveMatinsHymn's own doc comment already claimed the octave "keeps
the feast's own proper hymn all week," but the code never implemented
it -- it only matched when the day's own winner literally was
`assumption` (Aug 15 itself). Every other octave day (16, 17, 19, 20,
21, none of whom has a Matins hymn of their own authored) fell straight
through to the plain ferial hymn instead.
Adds a real octave-fallback tier between the existing per-feast override
and the seasonal tier. A deliberate departure from the reference engine,
which doesn't do this either (Monastic 1617's octave days have no
[Hymnus Matutinum] override at all) -- consistent with how this project
already treats octave readings/commemorations more generously than any
one source track.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VZSAgRi4QE4XRTqVto93zA
Fixed Aug 22 collided outright with the Assumption's own octave-closing
day (both this project's blended calendar tracks are kept deliberately,
so one shouldn't permanently suppress the other). Moved IHM to the
Saturday after the Feast of the Sacred Heart -- its real diocesan date
from 1914 until Pius XII's 1944 fixed-date decree, and also the date the
1969 reform returned to.
Required real new mechanism, not just a data move: a new bespoke
calendar/index.ts override (applyImmaculateHeart, following the existing
applyMarianSaturday/applyChristTheKing precedent), a move from the
sanctoral saints store to the temporal-feasts store (different propers
lookup entirely), and a real bug fix in matins.ts's nocturnReadingIds,
which only ever picked up a *sanctoral* winner's own reading file.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01VZSAgRi4QE4XRTqVto93zA
sundayPsalmNocturn/sundayCanticleNocturn only ever rendered a
versicle's V. line, silently missing the responsory for Sunday Matins
and every Duplex+ weekday feast using matins-psalmody-overrides. Both
now share the versicleText helper introduced for the ferial nocturn
antiphons, which combines V./R. into one block.
Every non-Sunday, non-Duplex+ Matins nocturn rendered bare psalms with
no antiphon at all. Sourced each psalm's antiphon from Divinum
Officium's Monastic Matins psalter and regrouped to fit
psalter-distribution.yml's own already-established (and non-historical)
weekday psalm redistribution, since the source's own per-weekday
antiphon pairings don't survive that reslice intact.
Every octave-reading yml's source field held only the Latin
attribution copied verbatim from the breviary text; add the matching
English translation now that the field carries both languages.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L3wxvz3mPXiHxkPB5JpnmD
The heading above each Matins octave reading was a bare Latin string
(e.g. "Sermo sancti Bernárdi Abbátis") rendered as-is in both the
Latin and English columns, since the app shows both languages side
by side rather than toggling. Widen OctaveReadingText.source and the
lesson part's label to carry per-language text, and render it as its
own lang-columns row like the body text beneath it.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L3wxvz3mPXiHxkPB5JpnmD
"Surge! Jam terris fera bruma cessit" / "Arise! the cold blasts from
earth have receded", read from the reference engine's Sancti/08-15.txt,
used throughout the Assumption's octave via that override mechanism.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L3wxvz3mPXiHxkPB5JpnmD
Matins always rendered the plain ferial hymn, unlike Lauds/Vespers which
already check a duplex-majus+ override then a seasonal default first.
Surfaced by the Assumption's octave, which keeps the feast's own proper
hymn all week rather than falling back to ferial.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L3wxvz3mPXiHxkPB5JpnmD
Assembled from already-verified Assumption pieces (Benedictus
antiphon, Vespers V/R, post-1950-dogma collect) since the octave
commemoration block this app's own generous-commemorations design
calls for has no direct reference-engine source to copy.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L3wxvz3mPXiHxkPB5JpnmD
The substitutedOctaveId gate only checked temporalCategory, not who
actually won the day, so a saint winning outright within an active
octave (e.g. St. Bernard vs. the Assumption octave, Aug 20) silently
lost that octave's own commemoration block at Lauds/Vespers. Now
mirrors resolveOfficeWinner's real gate (day.winner.kind and
ALWAYS_OVERRIDE_TEMPORAL_IDS).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L3wxvz3mPXiHxkPB5JpnmD
Monastic 1617 itself is inconsistent: the feast/closing octave day
render the old pre-dogma "Famulórum tuórum..." collect, but the
interior octave days (16/18/19/21) render the newer "Omnípotens
sempitérne Deus, qui Immaculátam..." collect from the 1954 Divino
Afflatu revision -- a data cross-reference slip in the reference
engine, not a deliberate rubric. vu already used one collect uniformly
across the whole octave rather than replicating the split; switch that
shared collect to the newer text per user preference.
Lauds/Vespers were falling back to a bare collect for her commemoration
since only the collect had been authored. Live-verified the antiphon
and versicle against the reference engine (Divino Afflatu - 1954,
1955-08-21, the nearest year her feast actually fell on a Sunday and
was commemorated rather than won).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L3wxvz3mPXiHxkPB5JpnmD
Two new mechanism-gap items: an arbitrary-date calendar picker (day-nav
currently only steps one day at a time) and a "read it to me" audio
interface via the Web Speech API for long texts.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L3wxvz3mPXiHxkPB5JpnmD
Visiting the app with no hour in the URL previously always showed
Prime for today, since selectedHour was hardcoded in AppState. Now it
picks the hour whose band contains the visitor's local clock time.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01L3wxvz3mPXiHxkPB5JpnmD
HOUR_IDS drove the nav's render order but was declared arbitrarily;
reorder to Matins, Lauds, Prime, Terce, Sext, None, Vespers, Compline
so the nav reads left-to-right the way the office actually unfolds.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Matches the reference app's row layout instead of a vertical sidebar
list, and switches from buttons to anchors since hours are real
routes — plain click still SPA-navigates, modifier/middle clicks
fall through to native browser behavior (open in new tab, etc.).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Matins is hand-assembled in TS rather than driven by a data/hours/*.yml
parts list, so it never got the closing wired in that every other hour
has. Live-verified against the reference engine's Monastic Tridentinum
1617 Matutinum "Conclusio" block that Matins' closing text is
byte-identical to lauds-conclusio — reusing it the same way
Vespers already does.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Same `wins: duplex` override as the Assumption's and Pentecost's
octaves, applied to two more common-class octaves. Accepted side
effect: St. Damasus (Dec 11) and St. Nicholas of Tolentino (Sep 10),
both Semiduplex, no longer tie-and-win an ordinary octave day here --
they're now commemorated instead of winning outright.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Demote St. John Eudes (Aug 19) from Duplex to Semiduplex -- too minor
a confessor to keep outranking the Assumption's octave. To make that
demotion actually cede the day, raise the Assumption's octave to
`wins: duplex` (same pattern already used by Pentecost's octave), so an
ordinary-day tie no longer automatically favors the occurring saint.
Promote St. Thomas of Canterbury (Dec 29) to Duplex so he keeps
winning against the Christmas octave stack now that ordinary-day ties
are no longer a given.
Also add a closing-day tie-break to octave-vs-octave precedence
(pickWinningOctave): when two active octaves tie in rank, the one on
its own closing day now wins the label contest, ahead of the existing
"more recently started" tie-break. Needed because Assumption's day 3
and St. Lawrence's own closing day (Aug 17) now tie at Duplex, and the
closing day should still govern that date's label.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
A different, wider rule than the psalm/canticle Gloria (Triduum-only):
responsories drop their Gloria for the whole of Passiontide (Passion
Sunday through Holy Saturday), except on a Sancti feast's own day —
live-verified (St. Joseph, Duplex I, 2027-03-19, landing in Passiontide
that year, keeps it; an ordinary ferial day in the same window omits it,
at both Compline and Prime).
Every responsory's stored text differs, Gloria-bearing vs Gloria-free,
by exactly one fixed-wording line, so resolve-common.ts's new
resolveResponsory filters that line out rather than requiring separate
per-season files or a data migration — mechanism first, per the
project's own stated content-authoring order. Wired into the four hours
that have a real responsory (Compline, Prime, Lauds, Vespers); Lauds/
Vespers' existing hand-authored Passiontide-proper responsories were
already Gloria-free and are unaffected.
Tenebrae's rubric replaces Matins' usual opening with a silently-said
Pater/Ave/Credo, live-verified against both Monastic Tridentinum 1617
and Divino Afflatu 1954 (Holy Thursday: Invitatorium and Hymnus both
"omittitur", no Ps 3 or versicle either). Since this app already never
renders that silent block on ordinary days either (data/hours/prime.yml),
the fix is a pure omission straight into Nocturn 1, using the same
isInTriduum window added for the Gloria Patri work.
Benedictus, Magnificat, and Nunc Dimittis had the Gloria Patri baked
verbatim into their propers YAML with no Triduum suppression; Matins'
Nocturn 3 OT canticles had none at all. Give the 'canticle' ResolvedPart
the same centrally-attached gloriaPatri field psalms already got, strip
the baked-in text, and drop Lauds' now-redundant local append/Triduum
check — live-verified that canticles follow the exact same Triduum-only
omission as psalms (e.g. Benedictus at Lauds of Holy Thursday).
Psalms across every hour rendered with no closing Gloria Patri at all.
Add it centrally (hours/index.ts) rather than per-hour, using a new
calendar/temporal.ts isInTriduum helper so it's correctly dropped from
Maundy Thursday through Holy Saturday and nowhere else — live-verified
against both of this app's actual source tracks (Divino Afflatu 1954
and Tridentine 1906), not just the modern Rubrics 1960 default. Also
fixes Lauds' existing but Triduum-blind canticle Gloria Patri append.
Adds a section-heading ResolvedPart kind and emits "Invitatory" and
"Nocturn I/II/III" headings in matins.ts, so the hour's structural
boundaries are visible instead of rendering as one undifferentiated
stream of parts. Ferial (1-nocturn) days get no Nocturn heading, matching
their existing single-unnumbered-nocturn design.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Builds every day label as a {en, la} pair (weekdays, ranks, season/ordinal
phrasing authored in Latin; proper names without an authored Latin form
fall back to their English string) and merges the date+day-label into one
header in the day-nav bar.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Nocturn 1 and 2 each get exactly one reading; Nocturn 3 absorbs the
rest of the pool, however large. A pool of only one reading goes to
Nocturn 3 rather than Nocturn 1.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Keeps Simplex commemorating in place, but a Semiduplex feast on an
ordinary Sunday goes back to transferring off to the next open day —
this app's own design choice, not a correction against a reference
engine. The transfer mechanism is already generic: this file's `decideOccurrence`
only signals "couldn't hold onto this candidate" for the day itself;
calendar/index.ts's resolveDay is what independently checks each
neighboring day's own native occurrence for an incoming transfer and
claims it — no separate lookback logic needed.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
At Vespers/Compline, a day's own Vigil losing tonight to tomorrow's First
Vespers was showing up as a redundant extra commemoration alongside that
same feast (e.g. Nov 29's Compline showing both "St. Andrew" and "Vigil
of St. Andrew" for Advent I's anticipated evening).
Added an explicit `vigilOf` field to each Vigil's own saint record
(several Vigil ids don't map cleanly to their feast's id by stripping a
`vigil-of-` prefix — `vigil-of-the-assumption` vs `assumption`,
`vigil-of-st-james` vs `st-james-the-greater` — so this needs to be
explicit rather than guessed from the id or date adjacency).
calendar/vespers.ts's evening-anticipation merge now skips adding the
Vigil commemoration when tomorrow's winner or its own commemorations
already include that same feast.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Two related fixes surfaced while chasing why St. Gregory Thaumaturgus
appeared to be misdated on Nov 18:
- calendar/commemorations.ts's `ordinary-sunday` case was transferring a
Semiduplex (or lower) saint off the Sunday instead of commemorating it
in place — an earlier, unverified guess. Live-verified against both
Tridentine 1906 and Divino Afflatu 1954 (St. Gregory Thaumaturgus, St.
Clement, St. Apollinaris, St. Thomas Becket, all real cases): the saint
stays and is commemorated on the Sunday itself, same as Simplex, never
pushed to the next open day. This is what was actually moving Gregory
onto Nov 18 — not a data error. `applyChristmasOctaveSunday`'s own
Dec 26-29 special case needed a matching adjustment (per its own
documented intent, the displaced saint there is deliberately *not*
commemorated in place, since he reappears in full on Dec 30 instead).
- getDayLabel had five branches, each hand-assembling its own
filter/format logic for which commemorations to show — which is why
the octave phrasing, the closing-day title, and a missing Sunday
commemoration turned into three separate bugs earlier instead of one.
Replaced with a single shared `collectCommemorations` used by every
branch. This also exposed that the sanctoral-winner branch never
showed any commemorated saint at all, and every other branch only
showed the *first* one (`.find()`), silently dropping real collisions
— both now show every commemorated saint, matching this app's own
generous-commemoration design.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Three related day-label bugs, all found via Aug 19's mislabeled Assumption
octave commemoration:
- A sanctoral winner that displaces the only active octave now shows the
octave's real "Nth Day within the Octave of X" phrasing (no rank, since
it's riding along under the winning feast), not a bare feast name.
- An octave's own closing day now titles itself "Octave of X", matching
the real DO's "in octava" vs. "infra octavam" distinction, instead of
"8th Day within the Octave of X".
- A sufficiently-ranked octave can now outrank a privileged temporal
season (e.g. the Immaculate Conception's octave outright winning several
of its days against Advent, live-verified against Divino Afflatu 1954),
via a new shared octaveGoverningPrivilegedDay helper used by both the
day label and the actual office-content resolver. Deliberately excludes
Christmastide, whose own stacked octaves are structurally already that
season's temporal content rather than a foreign add-on.
Updates a Matins test fixture that had unknowingly relied on the
Dec 15 bug (Advent ferial + a lone commemorated saint) and splits it into
a clean ferial case plus a dedicated commemorated-saint case.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Rank display (added in 1fc4bd7) only fired for a plain sanctoral
winner. Every other case -- an active octave day, a named temporal
feast, a season anchor day (Easter, Trinity Sunday, Epiphany), or the
plain ordinal Sunday/feria fallback -- showed no rank at all, which is
what made today's octave day (Aug 18, day 4 of the Assumption's
octave) show no rank. Octave days use the octave's own already-computed
effective rank; anchor days get a fixed Duplex I Class (Ash Wednesday
excepted -- a privileged feria has no duplex-scale rank); the plain
fallback derives a Semiduplex/Feria label from the day's own
temporalCategory, since Sundays/ferias aren't on the duplex scale at
all.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
TemporalFeastRecord had no rank field at all, so Christmas/Pentecost/
Christ the King/Marian Saturday had nothing for the day label to show.
Christmas, Pentecost, and Christ the King are Duplex I Class (the
latter's own file comment already said so); Marian Saturday is a
Simplex-strength votive commemoration.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Documents the Francis of Paola / Basil the Great live-verification,
the identical-apart-from-versicles finding, the Duplex+ saints checked
for proper-antiphon eligibility, and the M1617-kalendar wrinkle found
along the way (a SanctiM file existing doesn't mean the saint is
actually on the Monastic Tridentinum 1617 calendar).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Two new Common-tier proofs (Francis of Paola, Basil the Great) confirm
the resolver picks up both new categories with verified antiphon text.
The old ferial-fallback proof (St. John Bosco, Common of a Confessor)
is now superseded since that category is authored -- swapped for Ss.
Fabian and Sebastian (Common of Several Martyrs, still unauthored).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
will