temporal(ef): the Sacred Triduum's own Latin identity, RG 91 entry 2

Holy Thursday, Good Friday and Holy Saturday resolved to the generic
Passiontide ferial slugs with no distinct identity of their own -- rank
and colour were already correct (RG 91 entry 2; Holy Thursday's white
per RG 128(b)/RG 122), only Celebration.names/subject were missing.

Keeps the slug (RG 91 entry 2 is identified structurally, off rank and
Easter offset, by Precedence_ef.band -- never off the slug -- so
renaming it would touch precedence for no reason and break the lectio
differential's own slug mapping for these three lectionary keys, the
same reasoning already given for Holy Family/Holy Name). Carries the
identity in Celebration.names instead, the same treatment those two
established: Latin, not English, to avoid reading the oracle this axis
is compared against.

Names sourced from both photographic scans, word for word, corroborated
by the electronic transcription's own table-of-contents-style listing at
the identical three headings (three-way agreement, nothing to
adjudicate): "Feria V in Cena Domini" (Holy Thursday), "Feria VI in
Passione et Morte Domini" (Good Friday -- not the transcription's own
informal "Feria VI Parasceve" used in passing elsewhere), "Sabbato
Sancto" (Holy Saturday).

Also tags subject = Lord, going further than Holy Family/Holy Name
strictly required (they needed it only because band's entry 14 reads
subject to outrank a colliding fixed feast; entry 2 here needs no such
thing). Safe: disposition's RG 112(a) branch only fires when both sides
of an occurrence are tagged Lord, and no Lord-subject sanctoral entry
has a fixed civil date inside Holy Week's own movable range; any Class1
sanctoral entry landing there (e.g. a transferred Annunciation) reaches
disposition's earlier, subject-blind Transfer branch first regardless.
Verified against the full 1583-9999 domain: zero
observed/commemorations/transferred_* difference traceable to the tag.

test_golden.ml's describe gains a name_la field alongside subject, for
the same reason subject was added: a names-only change is otherwise
invisible to every pin whose rank/colour/slug are already correct (this
task's own catalogued vacuity flavour #1). Six existing pins move from
subject=temporal to subject=lord (the three Easter-extreme Holy Saturday
pins, 2016's Good Friday, and the 2026 Holy Thursday/Good Friday pair);
the rest gain name_la=- by construction (no other temporal-cycle
candidate carries a name yet).

Mutation-tested: reverting temporal_ef.ml alone (keeping the tests)
reddens the new Rite_ef unit test plus five golden pins across two
independent layers.

1 files changed, 81 insertions, 0 deletions

diff --git a/test/test_temporal_ef.ml b/test/test_temporal_ef.ml
index f341533..a979700 100644
--- a/test/test_temporal_ef.ml
+++ b/test/test_temporal_ef.ml
@@ -732,6 +732,85 @@ let test_holy_name_fallback_anchor_present_and_erosion_is_caught () =
   Alcotest.(check string) "sanity: temporal itself is unaffected by the anchors-list erosion"
     expected_slug (slug_of expected_date)
 
+(* Gap 1 (ef-triduum-litanies, docs/research/rules-register.md, "the Sacred
+   Triduum has no proper office"): the Missal's own Mass-propers headings for
+   Holy Thursday/Good Friday/Holy Saturday, both photographic scans, word for
+   word (missale-romanum-1962.pdf and "Missale Romanum 1962_text.pdf",
+   corroborated by the electronic transcription's own table-of-contents-style
+   listing at the identical three headings -- three-way agreement, no
+   scan-vs-transcription conflict to adjudicate): "FERIA QUINTA IN CENA
+   DOMINI" / running header "Feria V in Cena Domini"; "FERIA SEXTA / IN
+   PASSIONE ET MORTE DOMINI"; "SABBATO SANCTO". The slug stays the generic
+   Passiontide ferial one (RG 91 entry 2 already identifies the Triduum
+   structurally, off the day's own Easter offset and rank --
+   {!Rite_ef.Precedence_ef.band}'s own entry-2 branch -- never off the slug,
+   so renaming it would touch precedence for no reason and would break the
+   lectio differential's own slug mapping for these three lectionary keys);
+   the missing IDENTITY is carried in [names], the same treatment RG
+   17(b)/(a) (Holy Family/Holy Name) already established -- Latin, not
+   English, for the same zero-circularity reason those two entries give
+   (would otherwise read the oracle this exact axis is compared against).
+
+   [subject] is ALSO set to [Lord] here, going further than Holy Family/Holy
+   Name's own precedent strictly required (they needed it because
+   {!Rite_ef.Precedence_ef.band}'s entry 14 reads [subject] to outrank a
+   colliding fixed feast). The Triduum has no such live need: entry 2 is
+   decided purely by [rank]/Easter-offset, never [subject] -- confirmed by
+   reading {!Rite_ef.Precedence_ef.band}'s own entry-2 branch, which tests
+   neither. Tagged anyway because it is simply true (the Last Supper, the
+   Passion and Death, the entombment are textbook mysteries of the Lord) and
+   safe: {!Rite_ef.Precedence_ef.disposition}'s own RG 112(a) branch (Lord vs
+   Lord) only fires when BOTH sides are tagged [Lord], and no [Lord]-subject
+   sanctoral entry anywhere in data/ef/sanctoral.sexp has a FIXED civil date
+   inside Holy Week's own movable range (earliest 19 March, latest 24 April
+   -- the register's own subject audit lists exactly six [Lord]-tagged
+   entries, all fixed in January, February, August, September or November);
+   any [Class1] sanctoral entry that DOES land there (e.g. a transferred
+   Annunciation) reaches {!Rite_ef.Precedence_ef.disposition}'s EARLIER,
+   subject-blind "I class, not a Sunday -> Transfer" branch first, so RG
+   112(a)'s own branch is never reached for it regardless of this tag.
+   Measured empirically too: this task's own full 1583-9999 blast-radius
+   sweep shows no [observed]/[commemorations]/[transferred_*] difference
+   traceable to this tag anywhere in the domain. *)
+let test_sacred_triduum_identity () =
+  (* 2026: Easter 5 April, so Holy Thursday/Good Friday/Holy Saturday are
+     2/3/4 April. *)
+  Alcotest.(check string) "2026-04-02 Holy Thursday: slug stays the generic ferial key"
+    "ef-passiontide-2-thursday" (slug_of (d 2026 4 2));
+  Alcotest.(check bool) "2026-04-02 Holy Thursday: subject Lord" true
+    (subject_of (d 2026 4 2) = Sub.Lord);
+  Alcotest.(check (option string)) "2026-04-02 Holy Thursday: Latin name, both photographic scans verbatim"
+    (Some "Feria V in Cena Domini") (name_la_of (d 2026 4 2));
+  Alcotest.(check string) "2026-04-03 Good Friday: slug stays the generic ferial key"
+    "ef-passiontide-2-friday" (slug_of (d 2026 4 3));
+  Alcotest.(check bool) "2026-04-03 Good Friday: subject Lord" true
+    (subject_of (d 2026 4 3) = Sub.Lord);
+  Alcotest.(check (option string)) "2026-04-03 Good Friday: Latin name, both photographic scans verbatim"
+    (Some "Feria VI in Passione et Morte Domini") (name_la_of (d 2026 4 3));
+  Alcotest.(check string) "2026-04-04 Holy Saturday: slug stays the generic ferial key"
+    "ef-passiontide-2-saturday" (slug_of (d 2026 4 4));
+  Alcotest.(check bool) "2026-04-04 Holy Saturday: subject Lord" true
+    (subject_of (d 2026 4 4) = Sub.Lord);
+  Alcotest.(check (option string)) "2026-04-04 Holy Saturday: Latin name, both photographic scans verbatim"
+    (Some "Sabbato Sancto") (name_la_of (d 2026 4 4));
+  (* Neighbouring days carry no name/subject override -- the tag is exactly
+     three days wide, not the whole Passiontide-2 week. *)
+  Alcotest.(check (option string)) "2026-04-01 (Wednesday of Holy Week): no Latin name"
+    None (name_la_of (d 2026 4 1));
+  Alcotest.(check bool) "2026-04-01 (Wednesday of Holy Week): subject stays temporal" true
+    (subject_of (d 2026 4 1) = Sub.Temporal);
+  Alcotest.(check (option string)) "2026-04-05 Easter Sunday: no Latin name from this branch"
+    None (name_la_of (d 2026 4 5));
+  (* A second, independent year (1666: the latest possible Easter, 25 April,
+     already this file's own Easter-extreme witness elsewhere) -- proves the
+     mechanism generalises off [days_between easter d], not hand-pinned to
+     2026's own civil dates. Holy Thursday/Good Friday/Holy Saturday are
+     22/23/24 April 1666. *)
+  Alcotest.(check (option string)) "1666-04-22 Holy Thursday (latest-Easter year): Latin name still fires"
+    (Some "Feria V in Cena Domini") (name_la_of (d 1666 4 22));
+  Alcotest.(check bool) "1666-04-22 Holy Thursday: subject Lord" true
+    (subject_of (d 1666 4 22) = Sub.Lord)
+
 let test_totality () =
   (* Every day of 2026 yields an office without raising. Not a slug
      re-validation -- Slug.t is a private string validated on construction,
@@ -765,6 +844,8 @@ let suite_extra =
     Alcotest.test_case "ember days" `Quick test_ember_days;
     Alcotest.test_case "rogations" `Quick test_rogations;
     Alcotest.test_case "colours" `Quick test_colours;
+    Alcotest.test_case "sacred triduum identity (RG 91 entry 2, Gap 1)" `Quick
+      test_sacred_triduum_identity;
     Alcotest.test_case "christmastide feria slugs" `Quick test_christmastide_feria_slugs;
     Alcotest.test_case "named days carry their week" `Quick test_named_days_carry_their_week;
     Alcotest.test_case "totality" `Quick test_totality;