kernel+ef: fix round 1 -- lectionary caller-supplied, not eager

Critical (coordinator review): a clean `dune build` produced a `colitur`
that died at startup on EVERY subcommand, including ones touching no
lectionary data at all. Root cause was two-fold: data/ef/lectionary.sexp was
never added to the root default-build alias (only materialised as a side
effect of the test suite's own deps, which is why every check in the prior
report passed), and Rite_ef.context loaded it as a module-init side effect
via failwith, undoing Lectionary.load's own "never raises" promise at a
point no caller could catch.

Fixed structurally: Rite_ef.context is now a function taking ~lectionary,
Lectionary_ef.readings takes ~lectionary, and neither touches the filesystem
any more -- the same caller-supplied discipline the sanctoral layer already
had, restoring rite_ef.mli's own pre-existing claim about it and leaving a
seam for a future diocesan lectionary overlay. bin/main.ml grows
load_ef_lectionary, a sibling of load_ef_layer, routed through the same
colitur: %s / exit 2 path. data/ef/lectionary.sexp added to the root default
alias. Every caller of Rite_ef.context updated to supply it.

Also: two new tests that genuinely distinguish chain step 1 from step 2
(19 March 2026, Joseph's own proper over a competing temporal entry; 13
January 2030, Holy Family reached only through the temporal slug, the
Baptism entirely absent) -- the prior two tests both survived swapping the
chain order. Both new pins verified directly against the real data. The
chain's own comment now states plainly that its warrant is lectio's observed
behaviour, not a confirmed Missal citation, per the rules register's own
open item.

1 files changed, 15 insertions, 29 deletions

diff --git a/lib/rites/rite_ef/lectionary_ef.ml b/lib/rites/rite_ef/lectionary_ef.ml
index f9ce9ab..ec1aa0a 100644
--- a/lib/rites/rite_ef/lectionary_ef.ml
+++ b/lib/rites/rite_ef/lectionary_ef.ml
@@ -1,37 +1,23 @@
 open Colitur_kernel
 
-(* [data/ef/lectionary.sexp] is located relative to the BUILD TREE, not the
-   process's own cwd -- the same reasoning bin/main.ml's own [data_dir]
-   documents at length: cwd varies with how the binary that eventually links
-   this library is invoked (a user's shell for `dune exec colitur --`, a
-   dune cram test's own sandboxed temp directory, `_build/default/test/` for
-   `dune test`), and nothing in this project's build pins it to the
-   repository root. [Sys.executable_name] resolves through /proc/self/exe on
-   Linux, so it reports the executable's own canonical absolute path even
-   when launched through a symlink (dune's cram sandbox places exactly one;
-   confirmed working there already by bin/main.ml). Every executable that
-   links this library -- bin/main.exe, test/test_colitur.exe,
-   tools/*.exe -- sits exactly one directory below _build/default/, so
-   climbing up twice and back down into data/ef always finds the file,
-   regardless of the caller's own cwd. *)
-let data_path () =
-  Filename.dirname (Filename.dirname Sys.executable_name) ^ "/data/ef/lectionary.sexp"
-
-(* Loaded once, at module initialisation: the data is year-independent and
-   build_day is called 3 074 246 times over the full domain. *)
-let lectionary =
-  match Lectionary.load (data_path ()) with
-  | Ok l -> l
-  | Error e -> failwith ("lectionary_ef: " ^ e)
-
 (* Step 1: the observed celebration's own proper.
    Step 2: the day's own temporal slug.
 
-   Nothing here encodes "Lent has daily propers": the presence of an entry is
-   the discriminator. Verified against lectio -- Lent 1 Monday returns its own
-   Ezech 34:11-16, while Advent, Christmas and post-Pentecost Mondays return
-   their Sunday's Mass. *)
-let readings ~observed ~temporal ~date:_ ~temporal_at:_ =
+   Nothing here encodes "Lent has daily propers": the presence of an entry in
+   [lectionary] is the sole discriminator -- this function does not branch on
+   season, rank, or any other field to decide whether a temporal slug "ought"
+   to have its own Mass.
+
+   The WARRANT for that shape is lectio's own observed behaviour only, not a
+   confirmed Missal citation: Lent 1 Monday returns its own Ezech 34:11-16,
+   while Advent, Christmas and post-Pentecost Mondays return their Sunday's
+   Mass, in both streams. docs/research/rules-register.md records this
+   openly as unconfirmed against the primary source ("EF reading-selection
+   rules ... Have lectio's behaviour; confirm against the Missal's
+   ferial-Mass rubrics when coding") -- that confirmation has not been done;
+   do not read this comment as citing RG/the Missal for the SELECTION rule
+   itself, only [Lectionary.find]'s presence-or-absence as the mechanism. *)
+let readings ~lectionary ~observed ~temporal ~date:_ ~temporal_at:_ =
   match observed.Celebration.citations with
   | _ :: _ as cs -> cs
   | [] -> (