Add 4 more fixed Lord's/BVM feasts, and fix Our Lady of the Snows' Nocturn 3

Finding of the Holy Cross (May 3), the Transfiguration (Aug 6), the
Exaltation of the Holy Cross (Sep 14), and Dedication of the Basilica of
the Most Holy Savior (Nov 9, the Lateran) were all excluded under the
same stale "fixed feast of the Lord, not modeled" note that turned out
wrong for Our Lady of the Snows. All four have real, complete content in
the reference engine. Presentation of the BVM (Nov 21) was excluded
under the equivalent "fixed BVM feast" note and is added the same way.

- finding-of-the-holy-cross: fully proprium, own collect/antiphon/Matins
  throughout; now correctly outranks
  ss-alexander-eventius-theodulus-and-juvenal, matching what that
  saint's own file already documented as the live reality.
- transfiguration: same proprium shape. Its subordinate saints (Ss.
  Xystus/Felicissimus/Agapitus) stay excluded -- blocked by the real,
  separate octave/subordination-rendering gap, not by an unmodeled
  Transfiguration.
- exaltation-of-the-holy-cross: reuses the Finding of the Holy Cross's
  content wholesale (ex Sancti/05-03) -- collect/antiphon via a cross-id
  `propers` pointer, Matins nocturn-readings duplicated (no cross-id
  lookup exists for that mechanism).
- dedication-of-the-basilica-of-the-most-holy-savior: reuses the
  existing Common-of-a-Dedication (collect-c8, benedictus-antiphon-
  common-of-a-dedication), real unique Nocturn 2 vita of her own. Her
  secondary, St. Theodore, stays excluded for the same
  octave/subordination reason as above.
- presentation-of-the-bvm: real unique collect of her own; her
  Benedictus antiphon is a substituted Common reference
  (truncate-and-add-alleluia edit specific to this feast), authored as
  her own file.

Also: our-lady-of-the-snows.yml was missing its own Nocturn 3 (she
defers to Commune/C11's own Gospel-incipit-plus-Bede-homily, same as
Presentation of the BVM) -- added.

Test updates (not regressions): three tests had fixed proof dates
colliding with the newly-added feasts, updated to reflect the new,
correct behavior. npm test (365 tests) and tsc --noEmit both pass.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

src/data/calendar/saints/dedication-of-the-basilica-of-the-most-holy-savior.yml

@@ -0,0 +1,29 @@
+# Read directly from the reference engine (`../divinum-officium-reference`,
+# per memory vu-reference-engine-file-layout, not a live query): untagged/
+# default [Rank] block of `web/www/horas/Latin/Sancti/11-09.txt` -- "In
+# Dedicatione Archibasilicæ Ss. Salvatoris;;Duplex II. classis;;5;;ex C8".
+# Previously excluded entirely under a stale "fixed feast of the Lord,
+# not modeled" note in sanctoral-calendar.yml's own November comment --
+# wrong, same class of error as our-lady-of-the-snows.yml: she has real,
+# unique Lectio1-6(+6a) content of her own (the history of the Lateran
+# basilica's dedication under Pope St. Sylvester, through its later
+# restorations), and her `ex C8` Rule just means her collect/antiphon/
+# Nocturn-3-Gospel-and-homily come from the existing Common of a
+# Dedication (`collect-c8.yml`, `benedictus-antiphon-common-of-a-
+# dedication.yml` -- already authored, reused directly here, not
+# duplicated) -- the same Common already used by
+# dedication-of-the-basilicas-of-ss-peter-and-paul.yml.
+id: dedication-of-the-basilica-of-the-most-holy-savior
+name: "Dedication of the Basilica of the Most Holy Savior (the Lateran)"
+rank: duplex-2-classis
+common: common-of-a-dedication
+propers: null
+collectCommon: collect-c8
+# collect-c8's own text has no {N} placeholder to substitute (it's about
+# the building, not a saint) -- collectName is still required by
+# getDayCollect's own lookup, but its value is never actually used.
+collectName:
+  la: []
+  en: []
+minorHoursCommon: common-of-a-dedication
+benedictusCommon: common-of-a-dedication

src/data/calendar/saints/exaltation-of-the-holy-cross.yml

@@ -0,0 +1,20 @@
+# Read directly from the reference engine (`../divinum-officium-reference`,
+# per memory vu-reference-engine-file-layout, not a live query): untagged/
+# default [Rank] block of `web/www/horas/Latin/Sancti/09-14.txt` -- "In
+# Exaltatione Sanctæ Crucis;;Duplex majus;;4;;ex Sancti/05-03". Previously
+# excluded entirely under a stale "fixed feast of the Lord, same
+# convention as the Finding of the Holy Cross" note in
+# sanctoral-calendar.yml's own September comment -- wrong, same class of
+# error, now fixed alongside it (see finding-of-the-holy-cross.yml). Her
+# own [Rule] is `ex Sancti/05-03` -- she reuses the Finding of the Holy
+# Cross's content wholesale, so `propers` points at that saint's own
+# collect/antiphon files directly rather than duplicating them (same
+# cross-id-reuse pattern as st-michael.yml's `apparition-of-st-michael`).
+# The Matins nocturn-readings mechanism has no equivalent cross-id
+# lookup, so exaltation-of-the-holy-cross.yml duplicates
+# finding-of-the-holy-cross.yml's own reading content instead.
+id: exaltation-of-the-holy-cross
+name: "The Exaltation of the Holy Cross"
+rank: duplex-majus
+common: proper-to-the-finding-of-the-holy-cross
+propers: "finding-of-the-holy-cross"

src/data/calendar/saints/finding-of-the-holy-cross.yml

@@ -0,0 +1,18 @@
+# Read directly from the reference engine (`../divinum-officium-reference`,
+# per memory vu-reference-engine-file-layout, not a live query): untagged/
+# default [Rank] block of `web/www/horas/Latin/Sancti/05-03.txt` --
+# "Inventione Sanctæ Crucis;;Duplex II classis;;5.2;;". Previously
+# excluded entirely under a stale "fixed feast of the Lord, not modeled"
+# note in sanctoral-calendar.yml's own May comment -- wrong, same class
+# of error as our-lady-of-the-snows.yml: her [Rule] says `proprium` -- a
+# fully proper feast with real, complete unique text throughout
+# (collect, antiphon, 8 Matins lessons -- the file itself has no
+# Lectio9 at all, an unusual but genuine structural feature, not an
+# omission on this app's part). `Sancti/09-14.txt`'s Exaltation of the
+# Holy Cross reuses this same content wholesale (`ex Sancti/05-03`) --
+# see exaltation-of-the-holy-cross.yml.
+id: finding-of-the-holy-cross
+name: "The Finding of the Holy Cross"
+rank: duplex-2-classis
+common: proper-to-the-finding-of-the-holy-cross
+propers: "finding-of-the-holy-cross"

src/data/calendar/saints/presentation-of-the-bvm.yml

@@ -0,0 +1,24 @@
+# Read directly from the reference engine (`../divinum-officium-reference`,
+# per memory vu-reference-engine-file-layout, not a live query): untagged/
+# default [Rank] block of `web/www/horas/Latin/Sancti/11-21.txt` -- "In
+# Præsentatione Beatæ Mariæ Virginis;;Duplex majus;;4;;ex C11". Previously
+# excluded entirely under a stale "fixed BVM feast, excluded per the
+# usual convention" note in sanctoral-calendar.yml's own November comment
+# -- wrong, same class of error as our-lady-of-the-snows.yml. Unlike her,
+# this feast DOES have a real, unique [Oratio] of its own (see
+# presentation-of-the-bvm-collect.yml) and Nocturn 2 vita content
+# (Lectio4-6, St. John Damascene + St. Ambrose on the Blessed Virgin).
+# Her [Ant 1] is a substituted Common text (`@Commune/C10:Ant 2_`, with a
+# truncate-and-add-alleluia edit specific to this feast -- see
+# presentation-of-the-bvm-antiphon.yml), effectively proper by the same
+# "own [Name]/edit makes shared-formula text proper" reasoning as
+# st-lucy-collect.yml, so authored via `propers` rather than a Common
+# fallback. No Lectio7-9 override in her own file (defers to Commune/
+# C11's own Nocturn 3, same as our-lady-of-the-snows.yml's — see that
+# file's own Nocturn 3 comment for why it's duplicated rather than
+# cross-referenced).
+id: presentation-of-the-bvm
+name: "The Presentation of the Blessed Virgin Mary"
+rank: duplex-majus
+common: common-of-the-bvm
+propers: "presentation-of-the-bvm"

src/data/calendar/saints/transfiguration.yml

@@ -0,0 +1,19 @@
+# Read directly from the reference engine (`../divinum-officium-reference`,
+# per memory vu-reference-engine-file-layout, not a live query): untagged/
+# default [Rank] block of `web/www/horas/Latin/Sancti/08-06.txt` -- "In
+# Transfiguratione Domini Nostri Jesu Christi;;Duplex II classis;;5.1;;".
+# Previously excluded entirely under a stale "fixed feast of the Lord,
+# not modeled" note in sanctoral-calendar.yml's own August comment --
+# wrong, same class of error as our-lady-of-the-snows.yml: her [Rule]
+# says `proper`/`Festum Domini` -- a fully proper feast with real,
+# complete unique text throughout (collect, antiphon, all 9 Matins
+# lessons), not a Common fallback at all. Three secondary saints (Ss.
+# Xystus/Felicissimus/Agapitus, Aug 6) are still excluded, subordinate to
+# this feast under a real, separate mechanism gap (octave/subordination
+# rendering isn't built for a non-octave "loses outright to a Duplex II
+# Classis feast of the Lord" case) -- not fixed by this addition alone.
+id: transfiguration
+name: "The Transfiguration of Our Lord Jesus Christ"
+rank: duplex-2-classis
+common: proper-to-the-transfiguration
+propers: "transfiguration"

src/data/calendar/sanctoral-calendar.yml

@@ -83,16 +83,20 @@ days:
   "04-28": [st-vitalis]
   "04-29": [st-robert, st-catherine-of-siena]
   "04-30": [st-peter-martyr]
-  # May — 5/12 of the pre-1955 pull. Same exclusions as before: fixed
-  # feasts of the Lord (Finding of the Holy Cross May 3), and movable
-  # ones (Ascension, Trinity Sunday, Corpus Christi — all skipped as
-  # primary, since they're temporal, not sanctoral). Two stubs this month
-  # (Ss. Alexander/Eventius/Theodulus/Juvenal, St. Pudentiana) are
-  # genuinely un-winnable in any year, not just an artifact of which year
-  # was checked — see their own files.
+  # May — 5/12 of the pre-1955 pull. Movable feasts (Ascension, Trinity
+  # Sunday, Corpus Christi) are skipped as primary, since they're
+  # temporal, not sanctoral. The Finding of the Holy Cross (May 3) was
+  # also excluded here at first as "a fixed feast of the Lord, not
+  # modeled" — wrong, corrected 2026-08 (see finding-of-the-holy-cross.yml)
+  # — she has real, complete content and now wins the day outright, which
+  # is exactly what her competitor Ss. Alexander/Eventius/Theodulus/
+  # Juvenal's own file already documented as the live-verified reality
+  # ("loses the day to the Finding of the Holy Cross... this isn't just
+  # an artifact of the specific year checked"). St. Pudentiana is the
+  # remaining genuinely un-winnable stub this month — see her own file.
   "05-01": [ss-philip-and-james]
   "05-02": [st-athanasius]
-  "05-03": [ss-alexander-eventius-theodulus-and-juvenal]
+  "05-03": [finding-of-the-holy-cross, ss-alexander-eventius-theodulus-and-juvenal]
   "05-04": [st-monica]
   "05-06": [st-john-before-the-latin-gate]
   "05-07": [st-stanislaus]
@@ -161,29 +165,32 @@ days:
   # their own files. A genuinely new wrinkle this month: several dates
   # have a real, named secondary saint permanently subordinate to a
   # fixed, non-movable *octave* day (Lawrence's own Aug 10-17 octave, the
-  # Assumption's Aug 15-22 octave) or to the fixed Transfiguration (Aug
-  # 6) -- octaves and fixed Lord's feasts aren't modeled by this app (a
-  # known, explicitly deferred gap), so unlike the "genuinely unwinnable
-  # stub" pattern used elsewhere (where both competitors ARE modeled and
-  # the engine can correctly show the subordination), these can't be
-  # represented without incorrectly letting them win outright. Excluded
-  # entirely rather than guessed at: Ss. Xystus/Felicissimus/Agapitus
-  # (Aug 6, under the Transfiguration), Ss. Tiburtius and Susanna (Aug
-  # 11), St. Clare (Aug 12 -- notable: a genuinely major saint, not an
-  # obscure one, caught by this same gap), Ss. Hippolytus and Cassian
-  # (Aug 13), St. Agapitus (Aug 18), Ss. Timothy/Hippolytus/Symphorian
-  # (Aug 22). Our Lady of the Snows (Aug 5) was also excluded here at
-  # first under the same "fixed BVM feast" reasoning -- wrong, corrected
-  # 2026-08: unlike the genuinely-unmodeled cases above, she has real,
-  # complete content (Matins readings + Common-of-the-BVM collect/
-  # antiphon) and this app already models three other fixed BVM feasts
-  # the same way (assumption, immaculate-conception, nativity-bvm) -- see
-  # our-lady-of-the-snows.yml's own header. Added below.
+  # Assumption's Aug 15-22 octave) -- octaves aren't modeled by this app
+  # (a known, explicitly deferred gap; see "Octave content" in TODO.md),
+  # so unlike the "genuinely unwinnable stub" pattern used elsewhere
+  # (where both competitors ARE modeled and the engine can correctly show
+  # the subordination), these can't be represented without incorrectly
+  # letting them win outright. Excluded entirely rather than guessed at:
+  # Ss. Tiburtius and Susanna (Aug 11), St. Clare (Aug 12 -- notable: a
+  # genuinely major saint, not an obscure one, caught by this same gap),
+  # Ss. Hippolytus and Cassian (Aug 13), St. Agapitus (Aug 18), Ss.
+  # Timothy/Hippolytus/Symphorian (Aug 22). Our Lady of the Snows (Aug 5)
+  # and the Transfiguration (Aug 6) were also excluded here at first,
+  # each under a stale "fixed BVM/Lord feast, not modeled" note -- wrong,
+  # corrected 2026-08: both have real, complete content, and this app
+  # already models three other fixed BVM feasts the same way (assumption,
+  # immaculate-conception, nativity-bvm) -- see our-lady-of-the-snows.yml
+  # and transfiguration.yml's own headers. Added below. Ss.
+  # Xystus/Felicissimus/Agapitus, formerly excluded here as subordinate
+  # to an unmodeled Transfiguration, are still excluded -- now blocked by
+  # the same real octave/subordination-rendering gap as St. Clare et al.
+  # above, not by the Transfiguration itself being unmodeled.
   "08-01": [st-peter-ad-vincula, ss-machabees]
   "08-02": [st-stephen-i]
   "08-03": [finding-of-st-stephen]
   "08-04": [st-dominic]
   "08-05": [our-lady-of-the-snows]
+  "08-06": [transfiguration]
   "08-07": [st-donatus]
   "08-08": [ss-cyriacus-largus-and-smaragdus]
   "08-09": [vigil-of-st-lawrence, st-romanus]
@@ -206,9 +213,11 @@ days:
   # St. Hadrian (Sep 8) is the one exception -- he's blocked by the
   # Nativity's own day, not its octave, and that day IS already a mapped
   # sanctoral entry, so the ordinary collision engine handles him
-  # correctly. The Exaltation of the Holy Cross (Sep 14) is excluded as a
-  # fixed feast of the Lord, same convention as the Finding of the Holy
-  # Cross. St. Nicholas of Tolentino (Sep 10) and St. Januarius (Sep 19)
+  # correctly. The Exaltation of the Holy Cross (Sep 14) was also
+  # excluded here at first, same "fixed feast of the Lord" reasoning as
+  # the Finding of the Holy Cross -- wrong, corrected 2026-08 alongside
+  # it (see exaltation-of-the-holy-cross.yml). St. Nicholas of Tolentino
+  # (Sep 10) and St. Januarius (Sep 19)
   # are both monastic-calendar-specific: the former overrides what the
   # base Tridentine table treats as a plain Nativity-octave continuation,
   # the latter is added outright, present in neither the base table nor
@@ -230,6 +239,7 @@ days:
   "09-01": [st-giles, ss-twelve-brothers]
   "09-08": [nativity-bvm, st-hadrian]
   "09-10": [st-nicholas-of-tolentino]
+  "09-14": [exaltation-of-the-holy-cross]
   "09-16": [ss-cornelius-and-cyprian, ss-euphemia-lucia-and-geminianus]
   "09-19": [st-januarius]
   "09-20": [vigil-of-st-matthew]
@@ -271,15 +281,20 @@ days:
   # named saint). Ss. Four Crowned Martyrs (Nov 8, the octave's own final
   # day) hit the same unmodeled-octave gap as several saints in August/
   # September. The Dedication of the Basilica of the Most Holy Savior
-  # (Nov 9, i.e. the Lateran) is a fixed feast of the Lord (named for the
-  # Savior directly), excluded the same way the Transfiguration was in
-  # August -- its secondary, St. Theodore, is excluded along with it for
-  # the same unmodeled-primary reason. Unlike that one, the Nov 18
-  # Dedication of the Basilicas of Ss. Peter and Paul is about the
-  # Apostles, not the Lord, so it's kept in scope, same as the
-  # Commemoration of St. Paul and the two Chair-of-Peter feasts. The
-  # Presentation of the BVM (Nov 21, itself a monastic-calendar-only
-  # addition) is a fixed BVM feast, excluded per the usual convention.
+  # (Nov 9, i.e. the Lateran) and the Presentation of the BVM (Nov 21,
+  # itself a monastic-calendar-only addition) were also excluded here at
+  # first, each under a stale "fixed Lord's/BVM feast, not modeled" note
+  # -- wrong, corrected 2026-08: both have real, complete content -- see
+  # dedication-of-the-basilica-of-the-most-holy-savior.yml and
+  # presentation-of-the-bvm.yml's own headers. Added below. Its
+  # secondary, St. Theodore, is still excluded -- now blocked by the same
+  # real octave/subordination-rendering gap as St. Clare et al. in
+  # August, not by the Dedication itself being unmodeled (her own
+  # embedded `[Commemoratio 5]` block even carries St. Theodore's real
+  # collect already, ready to use once that mechanism exists). The
+  # Nov 18 Dedication of the Basilicas of Ss. Peter and Paul was already
+  # kept in scope (it's about the Apostles, not the Lord), same as the
+  # Commemoration of St. Paul and the two Chair-of-Peter feasts.
   #
   # Three monastic-calendar-specific reassignments this month, the most
   # of any month so far: St. Emilian (Nov 12) displaces St. Martin I,
@@ -296,6 +311,7 @@ days:
   # multi-day skip) -- this app's existing transfer machinery handles it
   # correctly; see tests/calendar/november-sanctoral.test.ts.
   "11-01": [all-saints]
+  "11-09": [dedication-of-the-basilica-of-the-most-holy-savior]
   "11-10": [ss-tryphon-and-companions]
   "11-11": [st-martin-of-tours, st-menna]
   "11-12": [st-emilian]
@@ -304,6 +320,7 @@ days:
   "11-17": [st-gregory-thaumaturgus]
   "11-18": [dedication-of-the-basilicas-of-ss-peter-and-paul]
   "11-19": [st-odo, st-pontian]
+  "11-21": [presentation-of-the-bvm]
   "11-22": [st-cecilia]
   "11-23": [st-clement, st-felicitas]
   "11-24": [st-chrysogonus]