Merge branch 'ef-rebootstrap': re-bootstrap the sanctoral, refresh the differential

Re-bootstraps data/ef/sanctoral.sexp from lectio's corrected generator
(322 -> 327 entries), propagating fixes colitur had inherited wholesale:
15 entries wrongly marked commemoration-only that are really III-class
feasts, four entries missing entirely, six commemorations an electronic
transcription had hidden, and two class corrections.

38639 days change across 1583-9999, in three shapes with no remainder:
2245 where a III-class feast now correctly beats a class-4 feria, 35174
where one of five restored commemorations appears, and 1220 on 13
January. All five restored entries were verified against a photographic
scan -- real, correctly ranked, and confined to their own dates.

13 January is NOT fixed by this. It moves from one wrong answer to a
different one, because RG 112 and a Holy Family office are both unbuilt.
Recorded as its own allow-list entry with a citation and a count, rather
than left inside a season-boundary entry's cardinality pin, and
deliberately not pinned by a golden test.

Then refreshes the lectio differential fixture, which predated the seven
EF defects fixed in that project. Nine allow-list entries close -- each
was a real divergence adjudicated in colitur's favour, and each closes
because lectio was fixed, with its RG citation preserved. The list goes
from twelve entries to four; the remaining 540 differences are fully
attributed with none unexplained.

1 files changed, 106 insertions, 11 deletions

diff --git a/test/test_golden.ml b/test/test_golden.ml
index fb3cff4..5292427 100644
--- a/test/test_golden.ml
+++ b/test/test_golden.ml
@@ -163,16 +163,21 @@ let test_easter_extreme_1666 () =
    the SAME extreme as 1666, reached independently), chosen because it is a
    near-term year rather than a 17th-century one.
 
-   DELIBERATE EXCLUSION, not an oversight: register §6's F4 finding
-   confirms 2038-03-06 (+sts-felicitas-perpetua), 2038-03-08 (+john-of-god)
-   and 2038-03-09 (+frances-rome) are WRONG -- three sanctoral entries
-   bootstrapped from lectio carry a Commemoration_only status that should be
-   a real III-class Feast (a bootstrap-generator defect traced to
-   missalemeum's own commemoration-id ranks, not fixable without touching
-   lectio). Pinning those three dates as "correct" would freeze a known bug
-   into a regression test that then fights its own fix -- the report's own
-   warning against exactly this. Only the Easter-week days (temporal-cycle
-   territory, untouched by the March sanctoral bug) are pinned here. *)
+   ORIGINAL EXCLUSION, now RESOLVED (ef-rebootstrap, 2026-08-12): this
+   comment used to say 2038-03-06 (+sts-felicitas-perpetua), 2038-03-08
+   (+john-of-god) and 2038-03-09 (+frances-rome) were deliberately left
+   unpinned because they were WRONG -- three sanctoral entries bootstrapped
+   from lectio carried a `Commemoration_only` status that should have been a
+   real III-class Feast. That defect was in lectio's own generator, not
+   fixable without touching lectio; lectio's generator has SINCE been fixed
+   (source SHA-256 6a25e634... -> 1b303ef2...) and colitur re-bootstrapped
+   from it, so this specific exclusion reason no longer holds. These three
+   dates, plus three more of the same shape (2008-04-02/04/05), are now
+   pinned below in [test_iii_class_feast_wins_class4_feria] -- the deferred
+   pin this comment always pointed at, not a new finding. Kept as a
+   temporal-cycle-only test regardless (the two concerns -- Easter-extreme
+   arithmetic and a sanctoral occurrence rule -- stay separate pins even
+   though nothing forces that any more). *)
 let test_easter_extreme_2038_late_modern () =
   check ~msg:"2038-04-24 Holy Saturday: still Passiontide, violet (RG 128)" 2038 4 24
     "2038-04-24 saturday season=passiontide week=2 slug=ef-passiontide-2-saturday rank=class-1 colour=violet \
@@ -580,6 +585,92 @@ let test_purification_on_a_sunday_2025 () =
     "2025-02-02 sunday season=time-after-epiphany week=4 slug=purification-of-the-blessed-virgin-mary \
      rank=class-2 colour=white comms=[] in=- out=[]"
 
+(* ------------------------------------------------------------------ *)
+(* RG 91 entry 24 ("Festa III classis, in calendario Ecclesiae universae
+   inscripta") outranks entry 28 ("feriae IV classis") -- ef-rebootstrap
+   (2026-08-12). Fifteen data/ef/sanctoral.sexp entries, 6 March-5 April,
+   were wrongly bootstrapped `rank = commemoration` (Commemoration_only,
+   which RG 91's table has no row for at all, register §6.1's own Maurice/
+   Thomas finding) where the 1962 calendarium gives each a real class
+   number instead; corrected upstream in lectio's own generator and
+   re-bootstrapped here (source SHA-256 1b303ef2...). Only 7 of the 15 ever
+   land on a date that collides with a plain CLASS-4 feria somewhere in the
+   domain -- fact-checked by a full 1583-9999 blast-radius sweep finding
+   zero diff rows for the other 8 anywhere in the domain, not assumed.
+
+   CORRECTED (ef-rebootstrap fix round 1, F1): this comment previously gave
+   a single reason for the other 8's silence -- "always fall inside Lent/
+   Passiontide proper, RG 91 entry 22, which already outranks entry 24" --
+   WRONG for 3 of the 8, measured directly against the domain-wide season
+   breakdown per slug: `gabriel-the-archangel` (24 Mar) lands in
+   PASCHALTIDE 226 times, `john-damascene` (27 Mar) 846 times,
+   `john-of-capistrano` (28 Mar) 1108 times -- always at `class-1` there
+   (inside the Easter Octave, RG 91 entry 2/10), never a class-4 feria.
+   Entry 22 itself is also narrower than "Lent/Passiontide" suggests (its
+   own text excludes the Ember days). The TRUE, general shield, checked
+   against every (season, rank) pair these 8 slugs' dates ever land in
+   domain-wide: whichever office already occupies the date is EITHER
+   class-1 or class-2 (an Easter-Octave/Holy-Week/I-class-Sunday day,
+   entries 2/6/7/10; a II-class Sunday or Ember feria, entries 15/18) --
+   which trivially outranks ANY class-3 candidate regardless of status, and
+   also excludes it from commemoration by RG 111(a)/(b)'s own rank floors
+   the same way regardless of status -- OR the ordinary Lent/Passiontide
+   class-3 feria itself, entry 22, which outranks entry 24 for OBSERVATION
+   but (RG 111(c)/(d), no ferial-class gate on an "ordinary" sanctoral
+   candidate) admits the losing saint's commemoration identically whether
+   `Commemoration_only` or a genuine `Feast` -- so no context these 8 dates
+   ever reach anywhere in the domain distinguishes the two statuses in
+   either direction, observation or commemoration. Not "always Lent" --
+   "always something at entry ≤22 already occupies the day". These are the
+   SIX dates the task brief that started this re-bootstrap named as
+   producing "a wrong observed office"
+   under the OLD data (2008-04-02/04/05, 2038-03-06/08/09) --
+   [test_easter_extreme_2038_late_modern]'s own header above is the record
+   of why they were deliberately left unpinned until now. *)
+
+let test_iii_class_feast_wins_class4_feria () =
+  (* Two [omitted_has] checks, one per season family (Paschaltide,
+     Septuagesima), prove the displaced CLASS-4 feria actually reached
+     [disposition] and was actively omitted under RG26 ("IV-class ferias
+     are never commemorated"), not merely absent because no candidate was
+     ever offered -- this file's own established convention (the RG16(a)
+     and "IV-class ferias never commemorated" tests above), not repeated
+     for all six to avoid duplicating the same proof six times. *)
+  let d1 = fetch 2008 4 2 in
+  Alcotest.(check bool)
+    "2008-04-02: the displaced class-4 Paschaltide feria is in [omitted] (RG26), proving RG91 entry24>28 fired"
+    true
+    (omitted_has d1 "ef-easter-2-wednesday");
+  Alcotest.(check string) "2008-04-02: Francis of Paola (III class, RG91 entry 24) outranks the class-4 feria"
+    "2008-04-02 wednesday season=paschaltide week=2 slug=francis-of-paola rank=class-3 colour=white comms=[] \
+     in=- out=[]"
+    (describe d1);
+  check ~msg:"2008-04-04: Isidore of Seville (III class) outranks the class-4 Paschaltide feria, same shape" 2008
+    4 4
+    "2008-04-04 friday season=paschaltide week=2 slug=isidore-of-seville rank=class-3 colour=white comms=[] in=- \
+     out=[]";
+  check ~msg:"2008-04-05: Vincent Ferrer (III class) outranks the class-4 Paschaltide feria, same shape" 2008 4 5
+    "2008-04-05 saturday season=paschaltide week=2 slug=vincent-ferrer rank=class-3 colour=white comms=[] in=- \
+     out=[]";
+  let d4 = fetch 2038 3 6 in
+  Alcotest.(check bool)
+    "2038-03-06: the displaced class-4 Septuagesima feria is in [omitted] (RG26), proving RG91 entry24>28 fired"
+    true
+    (omitted_has d4 "ef-septuagesima-2-saturday");
+  Alcotest.(check string)
+    "2038-03-06: Sts Perpetua and Felicity (III class, RG91 entry 24) outrank the class-4 feria; RED per their \
+     own martyrs' colour, not the season's violet"
+    "2038-03-06 saturday season=septuagesima week=2 slug=sts-felicitas-perpetua rank=class-3 colour=red comms=[] \
+     in=- out=[]"
+    (describe d4);
+  check ~msg:"2038-03-08: John of God (III class) outranks the class-4 Septuagesima feria, same shape" 2038 3 8
+    "2038-03-08 monday season=septuagesima week=3 slug=john-of-god rank=class-3 colour=white comms=[] in=- \
+     out=[]";
+  check ~msg:"2038-03-09: Frances of Rome (III class) outranks the class-4 Septuagesima feria, same shape" 2038 3
+    9
+    "2038-03-09 tuesday season=septuagesima week=3 slug=frances-rome rank=class-3 colour=white comms=[] in=- \
+     out=[]"
+
 let suite =
   ( "golden pins (known-tricky years)",
     [ Alcotest.test_case "Easter extreme: 1598 earliest (22 Mar, Gauss-verified)" `Quick
@@ -612,5 +703,9 @@ let suite =
         `Quick test_rg16a_lord_feast_no_sunday_commemoration_2028;
       Alcotest.test_case
         "RG16(a), the Purification (oracle-ruled): observed outright, the Sunday not commemorated (2025-02-02)"
-        `Quick test_purification_on_a_sunday_2025
+        `Quick test_purification_on_a_sunday_2025;
+      Alcotest.test_case
+        "ef-rebootstrap: RG91 entry24 (III-class feast) outranks entry28 (class-4 feria) -- the six dates \
+         deferred since 2038's own Easter-extreme pin (2008-04-02/04/05, 2038-03-06/08/09)"
+        `Quick test_iii_class_feast_wins_class4_feria
     ] )