fix(ef): twelve RG 124 colours -- missalemeum has the rule inverted

RG 124, "De coloribus paramentorum", section C, both photographic scans word
for word: 124(b) gives red to "Sanctorum Apostolorum et Evangelistarum, in
eorum die natalicio, excepto festo S. Ioannis (27 decembris)", and 124(e) to
"Sanctorum Martyrum, quorum colitur aut martyrium". Red is for Apostles,
Evangelists and Martyrs. A Confessor, Doctor or non-martyr Virgin is white.

missalemeum has twelve of these INVERTED IN BOTH DIRECTIONS: ten confessors,
doctors and virgins in red (Ephrem, Julia of Falconieri, John Gualbert,
Camillus de Lellis, Jerome Emiliani, Martha, Alphonsus Liguori, Augustine,
Rose of Lima, John of San Fecundo), and two actual martyrs in white
(Apollinaris and Josaphat, both Bishop and Martyr). Both directions failing is
why this reads as one rule applied backwards rather than twelve separate
slips, and why they are corrected as a block.

Adjudicated by colitur against the scans -- its C18 carries the per-slug
reasoning and its M21 the oracle side, verdict colitur. Two of its fourteen,
conversion-of-st-paul and chair-of-st-peter, already agreed here.

Found by a full-year sweep comparing clectio's output against colitur day by
day, not by the patch tool that produced the earlier corrections. That tool
reads colitur's sanctoral.sexp directly and so misses everything in
adjustments.sexp, the overlay where all twenty of colitur's colour edits
actually live -- its SECOND structural blind spot, after temporal days.

All twelve go into the generator's missalOverrides table, so regeneration
preserves them, and into TestTridentineMissalOverrides, now sixteen pinned
fields. Mutation-tested in BOTH directions: reverting a martyr to white and a
confessor to red each redden the test with the rubric named. Pinning only one
direction would have let a careless "make them all white" fix pass ten of
twelve.

The oracle allow-list gains one entry covering the twelve dates in both years.

0 files changed, 0 insertions, 0 deletions