rite(ef): fix entry 8's Sunday exception and entry 14's over-restriction

Review of 436ba75 found two calendar defects and one coupling risk.

Entry 8 (All Souls) dropped register line 334's own qualifier,
"(yields to an occurring Sunday)" -- it returned 8 unconditionally,
so on 2 November falling on a Sunday (2025, 2031, 2036, 2042 in the
2005-2050 differential window), All Souls incorrectly outranked and
observed over the Sunday. 2 November is always Time_after_pentecost
and never coincides with any other entry's own date, so the only
rival this exception ever has is an ordinary entry-15 II-class
Sunday; on such a Sunday, band now returns one worse than entry 15's
own value rather than the literal 8 -- strictly worse, not merely
different, since an exact tie would fall to Precedence.resolve's slug
tie-break, which for "ef-all-souls" against a Sunday slug would make
All Souls win the tie anyway. Entry 8's own rank is untouched, so
Task 8's disposition still sees a true I-class candidate to transfer.

Entry 14 (Feasts of the Lord, II class) added a universal-layer
restriction the register's line 341 does not carry -- contrast entry
16 at line 342, which explicitly says "not of the Lord"; RG 37c
(register line 393) also speaks of "II-class feasts of the Lord"
with no universal qualifier. Dropped the layer test: a proper or
indult feast of the Lord now bands 14, not 19/20.

Also exposes vigil_suffix and ember_prefixes from precedence_ef.mli,
matching universal_layer/indult_prefix's existing "colitur convention,
not an RG citation" treatment -- both were previously private literals
duplicated in the test file, so a rename of temporal_ef.ml's slug
format could have drifted silently past both sides agreeing with each
other. Two test rows now build their candidate from Temporal_ef.temporal's
own real output (entry 18's Lent Ember day, entry 21's Ascension Vigil)
instead of a hand-typed slug, closing that specific coupling.

Adds the three cheap rows review named as closing the remaining
unpinned guards (a temporal-origin Class1 candidate on an unnamed
date; a universal Class2 vigil of the Lord; a non-universal Class3
vigil), plus a resolve()-level test proving the Sunday is actually
observed over All Souls, not just that band returns the right integer
in isolation.

2 files changed, 102 insertions, 36 deletions

diff --git a/lib/rites/rite_ef/precedence_ef.ml b/lib/rites/rite_ef/precedence_ef.ml
index 50da4a7..0716177 100644
--- a/lib/rites/rite_ef/precedence_ef.ml
+++ b/lib/rites/rite_ef/precedence_ef.ml
@@ -9,22 +9,27 @@
    without a separate exclusion for every later entry it pre-empts.
 
    Two kinds of evidence decide an entry:
-   - The temporal cycle's own office (Nativity, a Sunday, a feria, a vigil of
-     the Lord) is identified structurally, from the context's date/season/
-     weekday and the day's Easter offset -- never from its slug, which is
-     just a label. [origin = Temporal] gates every such entry so a sanctoral
-     candidate that happens to share a date (Immaculate Conception can never
-     coincide with the movable cycle, but nothing stops a future rite bug
-     from producing one) cannot be mistaken for the office itself.
-   - A sanctoral feast's entry (11-13 I class, 14/16/19/20 II class, 23/24
-     III class) is decided by its [rank] plus, per the brief's structural
-     insight, its {!Celebration.t}.layer: a celebration whose layer is not
-     the universal base is an overlay, hence "proper" or "indult" rather
-     than the universal entry (see precedence_ef.mli). [origin = Sanctoral]
-     gates these for the same reason: temporal-origin celebrations carry the
-     literal layer id "temporal" (rite_ef/temporal_ef.ml's [build]), which is
-     not [universal_layer] either, and would otherwise be misread as
-     "proper" by the layer test alone.
+   - The temporal cycle's own office (Nativity, a Sunday, a feria, All Souls)
+     is identified structurally, from the context's date/season/weekday and
+     the day's Easter offset -- never from its slug, which is just a label.
+     [origin = Temporal] gates every such entry so a sanctoral candidate that
+     happens to share a date (Immaculate Conception can never coincide with
+     the movable cycle, but nothing stops a future rite bug from producing
+     one) cannot be mistaken for the office itself. All Souls (entry 8, the
+     one non-temporal-origin member of this group) additionally reads the
+     context's weekday for its own register-stated exception -- see entry 8
+     below.
+   - A sanctoral feast's entry is decided by its [rank], and -- except at
+     entry 14 (see its own comment below, where the register draws no such
+     line) -- per the brief's structural insight, also by its
+     {!Celebration.t}.layer: a celebration whose layer is not the universal
+     base is an overlay, hence "proper" or "indult" rather than the
+     universal entry (11-13 I class; 16/19/20 II class; 23/24 III class; see
+     precedence_ef.mli). [origin = Sanctoral] gates these for the same
+     reason: temporal-origin celebrations carry the literal layer id
+     "temporal" (rite_ef/temporal_ef.ml's [build]), which is not
+     [universal_layer] either, and would otherwise be misread as "proper" by
+     the layer test alone.
 
    Vigils (21, 26) are the one shape neither of those two kinds fully
    describes on their own: a II/III-class vigil can be temporal-origin (the
@@ -46,19 +51,33 @@ let unclassified = max_int
 let is_indult layer = String.starts_with ~prefix:indult_prefix layer
 let is_universal layer = String.equal layer universal_layer
 
-(* Ember days are identified by the temporal cycle's own slug convention
-   (rite_ef/temporal_ef.ml's [ember]: "ef-<set>-ember-<day>"), not
-   re-derived here: the September anchor in particular is one of the more
-   contested dates in the 1962 calendar (temporal_ef.ml's own comment on
+(* Not an RG citation either -- see [universal_layer] above. Nothing in
+   {!Celebration.t} otherwise marks "this is a vigil, not an ordinary office
+   of the same rank" (see the file's top comment), so entries 21/26 read it
+   off the temporal cycle's own slug suffix (rite_ef/temporal_ef.ml's
+   [named], e.g. "ef-ascension-vigil"). Exposed so a future task naming a
+   sanctoral vigil (Task 10: Assumption, John Baptist, Peter & Paul,
+   Lawrence -- only Ascension exists today) uses the same suffix; a
+   differently-named vigil would band 16/24 instead of 21/26, silently. *)
+let vigil_suffix = "-vigil"
+
+let is_vigil slug = String.ends_with ~suffix:vigil_suffix slug
+
+(* Not an RG citation -- see [universal_layer]. Entry 18's Ember days are
+   identified by the temporal cycle's own slug convention (rite_ef/
+   temporal_ef.ml's [ember]: "ef-<set>-ember-<day>"), not re-derived here:
+   the September anchor in particular is one of the more contested dates in
+   the 1962 calendar (temporal_ef.ml's own comment on
    [third_sunday_of_september]), and re-deriving it a second time would only
    create a second place for that same uncertainty to drift. Only the
    Advent, Lent and September sets are listed: RG 91 entry 18 names exactly
    those three; the Whitsun (Pentecost) set is I class and falls inside the
-   Pentecost octave, entry 10, matched below before this is ever reached. *)
-let is_ember_18 slug =
-  String.starts_with ~prefix:"ef-advent-ember-" slug
-  || String.starts_with ~prefix:"ef-lent-ember-" slug
-  || String.starts_with ~prefix:"ef-september-ember-" slug
+   Pentecost octave, entry 10, matched below before this is ever reached.
+   Exposed for the same reason as [vigil_suffix]: a rename of temporal_ef's
+   format has somewhere to be caught other than a silently-wrong entry 18. *)
+let ember_prefixes = [ "ef-advent-ember-"; "ef-lent-ember-"; "ef-september-ember-" ]
+
+let is_ember_18 slug = List.exists (fun prefix -> String.starts_with ~prefix slug) ember_prefixes
 
 let band (ctx : Vocab_ef.season Precedence.context) (c : Vocab_ef.rank Precedence.candidate) :
     int =
@@ -68,7 +87,7 @@ let band (ctx : Vocab_ef.season Precedence.context) (c : Vocab_ef.rank Precedenc
   let layer = cel.Celebration.layer in
   let slug = Slug.to_string cel.Celebration.slug in
   let is_temporal = c.Precedence.origin = Precedence.Temporal in
-  let is_vigil = String.ends_with ~suffix:"-vigil" slug in
+  let is_vigil = is_vigil slug in
   let date = ctx.Precedence.date in
   let season = ctx.Precedence.season in
   let weekday = ctx.Precedence.weekday in
@@ -77,6 +96,11 @@ let band (ctx : Vocab_ef.season Precedence.context) (c : Vocab_ef.rank Precedenc
   (* Easter offset, the same convention as temporal_ef.ml's [days_between
      easter d]: 0 is Easter itself, negative before, positive after. *)
   let off = Date.to_rata date - Date.to_rata (Computus.gregorian_easter (Date.year date)) in
+  (* Named so entry 8's Sunday exception below can read "one worse than the
+     Sunday it must yield to" rather than a bare integer that happens to
+     equal entry 15's own value; entry 15's own branch returns this same
+     binding, not a second literal, so the two can never drift apart. *)
+  let entry_15_band = 15 in
   let open Vocab_ef in
   (* 1: Nativity, Easter Sunday, Pentecost Sunday (I class w/ octave). *)
   if is_temporal && rank = Class1 && ((m = 12 && d = 25) || off = 0 || off = 49) then 1
@@ -101,8 +125,23 @@ let band (ctx : Vocab_ef.season Precedence.context) (c : Vocab_ef.rank Precedenc
     (* 7: I-class ferias not above -- Ash Wednesday; Mon/Tue/Wed of Holy Week.
        Thu-Sat of Holy Week are the Triduum, entry 2 above, not this entry. *)
   else if is_temporal && rank = Class1 && (off = -46 || (off >= -6 && off <= -4)) then 7
-    (* 8: All Souls. *)
-  else if (not is_temporal) && rank = Class1 && m = 11 && d = 2 then 8
+    (* 8: All Souls -- register line 334's own text carries a qualifier this
+       transcription must honour: "yields to an occurring Sunday". 2 November
+       is always Time_after_pentecost (well clear of Advent/Lent/Passiontide
+       and of every other entry's own Easter-relative or fixed date), so a
+       Sunday landing on it is always an ordinary entry-15 II-class Sunday --
+       the one and only rival this exception ever has to lose to. On such a
+       Sunday this returns [entry_15_band + 1]: strictly worse than 15 (an
+       exact tie would fall to Precedence.resolve's slug tie-break, which
+       for "ef-all-souls" against a "ef-time-after-pentecost-sunday-*" slug
+       would make All Souls WIN -- the precise bug this guards against), but
+       otherwise not a citation to any other RG 91 row -- nothing else can
+       ever occur on 2 November to be confused with it. Entry 8's own [rank]
+       is untouched by this, so Task 8's disposition (RG 95: only I-class
+       feasts transfer) still sees the true I-class candidate it needs to
+       move to 3 November. *)
+  else if (not is_temporal) && rank = Class1 && m = 11 && d = 2 then
+    if is_sunday then entry_15_band + 1 else 8
     (* 9: Vigil of Pentecost. *)
   else if is_temporal && rank = Class1 && off = 48 then 9
     (* 10: Days within the Octaves of Easter and Pentecost. *)
@@ -116,12 +155,16 @@ let band (ctx : Vocab_ef.season Precedence.context) (c : Vocab_ef.rank Precedenc
        not the universal layer (11), and marked as an indult overlay (12's
        "not indult" test having just failed). *)
   else if (not is_temporal) && (not is_vigil) && rank = Class1 then 13
-    (* 14: Feasts of the Lord, II class. *)
-  else if (not is_temporal) && (not is_vigil) && rank = Class2 && is_universal layer
-          && subject = Subject.Lord
-  then 14
+    (* 14: Feasts of the Lord, II class -- register line 341, deliberately
+       UNQUALIFIED (contrast entry 16 at line 342, which explicitly says "not
+       of the Lord"; RG 37c, register line 393, speaks of "II-class feasts of
+       the Lord" replacing an occurring II-class Sunday with no universal
+       qualifier either). No layer test here, unlike 11/12/13 and 16/19/20:
+       the register does not split this entry into universal/proper/indult,
+       so a proper or indult feast of the Lord still bands 14, not 19/20. *)
+  else if (not is_temporal) && (not is_vigil) && rank = Class2 && subject = Subject.Lord then 14
     (* 15: Sundays, II class (every Sunday not already named at 6). *)
-  else if is_temporal && rank = Class2 && is_sunday then 15
+  else if is_temporal && rank = Class2 && is_sunday then entry_15_band
     (* 16: II-class feasts of the universal Church, not of the Lord. *)
   else if (not is_temporal) && (not is_vigil) && rank = Class2 && is_universal layer then 16
     (* 17: Days within the Octave of the Nativity (26-28 Dec are Stephen,
diff --git a/lib/rites/rite_ef/precedence_ef.mli b/lib/rites/rite_ef/precedence_ef.mli
index e8d4a11..63fb0a1 100644
--- a/lib/rites/rite_ef/precedence_ef.mli
+++ b/lib/rites/rite_ef/precedence_ef.mli
@@ -18,6 +18,24 @@ val universal_layer : string
 (** See {!universal_layer}. *)
 val indult_prefix : string
 
+(** Slug suffix marking a celebration as a vigil (RG 91 entries 21, 26),
+    e.g. "ef-ascension-vigil". Also colitur's own convention, not an RG
+    citation, exposed for the same reason as {!universal_layer}: only the
+    Ascension Vigil exists today (rite_ef/temporal_ef.ml); the Assumption,
+    John Baptist, Peter & Paul and Lawrence vigils arrive as sanctoral data
+    in a future task, and must use this same suffix or {!band} will band
+    them 16/24 (an ordinary feast of the same rank) instead of 21/26. *)
+val vigil_suffix : string
+
+(** Slug prefixes marking a celebration as one of RG 91 entry 18's three
+    Ember-day sets (Advent, Lent, September -- the Pentecost/Whitsun set is
+    I class and matched by entry 10 before this is ever consulted). Also
+    colitur's own convention mirroring rite_ef/temporal_ef.ml's own "ef-<set>
+    -ember-<day>" slug format, not re-derived from first principles; exposed
+    so a rename of that format has somewhere to be caught other than a
+    silently-wrong entry 18. *)
+val ember_prefixes : string list
+
 (** Returned for a candidate shape RG 91's 28-entry table has no row for --
     e.g. a [Class1] vigil that is not the Nativity or Pentecost (entries 5,
     9 are the only I-class vigils the table names), or a [Class4] candidate
@@ -29,7 +47,12 @@ val unclassified : int
 
 (** [band ctx c]: RG 91's Table of Precedence. Returns the table's own entry
     number -- I class 1-13, II class 14-21, III class 22-26, IV class 27-28;
-    lower wins (see {!Precedence.rules.band}). Total over every candidate
-    {!Precedence.resolve} or {!Calendar} can construct, including shapes the
-    1962 table itself does not describe (see {!unclassified}). *)
+    lower wins (see {!Precedence.rules.band}) -- EXCEPT where the table's own
+    text states an exception: entry 8 (All Souls, register line 334) reads
+    "yields to an occurring Sunday", so on a Sunday this returns a value that
+    loses to entry 15 rather than the literal integer 8 (see the comment on
+    entry 8 in precedence_ef.ml for the exact value and why). Total over
+    every candidate {!Precedence.resolve} or {!Calendar} can construct,
+    including shapes the 1962 table itself does not describe (see
+    {!unclassified}). *)
 val band : Vocab_ef.season Precedence.context -> Vocab_ef.rank Precedence.candidate -> int