| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| | |
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The Latin Mass Society Ordo 2024-2025 for England and Wales (v2.21,
compiled by Peter Day-Milne) gives 9 August 2025 as "OUR LADY on SATURDAY
IV Cl W / V Mass of BVM / Commem of St Romanus M at Low Mass" -- no vigil
-- and 10 August as the 9th Sunday after Pentecost with St Laurence
commemorated. That is colitur's output on both days, field for field.
It is discriminating where the Ordo already in docs/research was not. This
edition lists six vigils as the day's office across its year, including the
Assumption's on 14 August 2025, and its footnotes discuss arranging a
diocesan feast around the Vigil of St Laurence in other years. Its absence
on 9 August 2025 is an application of the rule, not a data gap. The
extraordinaryform.org Ordo omits that vigil on every date it covers,
including 2027-08-09 where the clause says to keep it, which is why it
witnesses nothing here.
Also confirms the Vigil of the Assumption as II Cl violet, the RG 128
colour this project corrected from the bootstrap's White.
The clause's other shape -- feast transferred away, 2022/2033/2038/2044 --
remains unwitnessed; no accessible Ordo covers those years. Stated in the
note rather than left implied.
|
| |\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The last unbuilt clause in the EF module. A II/III-class vigil is omitted
entirely when the feast it precedes fails to keep its own day -- 1 744 days
across 1583-9999, three verified shapes, zero unexplained.
Rule and vigil inventory independently confirmed by O'Connell 1964, outside
the Divinum Officium -> missalemeum -> lectio lineage. The 10 affected days
in 2005-2050 are allow-listed as C39; no external day-level witness exists,
which the note states rather than glosses.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Replaces the withdrawn Ordo claim with a real one of a different kind.
O'Connell, The Celebration of Mass (4th ed., 1964) -- a rubrics commentary
contemporary with the 1960 Code, outside the Divinum Officium ->
missalemeum -> lectio lineage, already cited here for RG 128 -- states all
three of RG 33's triggers in English, the third included, verbatim in
substance with the Latin.
The same passage confirms vigil_feast_table's membership independently:
Christmas Eve and Pentecost I class and so outside the rule; Ascension,
Assumption, John the Baptist and Peter & Paul II class; St Lawrence the
sole III-class vigil. Five entries, the same five, same class split.
It confirms the RULE and the ENTITY LIST, not the 10 days -- O'Connell
publishes no day-by-day calendar. The distinction is stated in the note
rather than blurred, since blurring it is what the withdrawn Ordo
paragraph did.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
C39 and the 2025 golden pin both claimed the published Ordo independently
confirms RG 33's third trigger, because it omits St Lawrence's vigil on
2025-08-09 where lectio and missalemeum keep it.
It does not. The Ordo omits that vigil on 2027-08-09 as well, where the
clause says to keep it (10 August is a Tuesday, the feast holds its day),
and 2026-08-09 is a Sunday where the rule's first trigger omits it anyway.
The 2024 date has no row. So the Ordo never shows St Lawrence's vigil on
any date it covers, and the 2025 agreement is a blanket gap coinciding with
the right answer, not evidence.
The rule itself is unaffected: it rests on RG 33's text, which admits no
latitude, and on lectio not evaluating the clause at all -- "RG 33" appears
nowhere in its production code, and the third trigger cannot emerge from
per-day rank comparison because it turns on what is observed the following
day. The divergence is architectural, not a disagreement about the rubric.
Recorded rather than deleted: the withdrawn claim is why this entry was
first written up as unusually well corroborated.
|
| | |
| |
| |
| |
| |
| |
| | |
The note apportioned the entry's ten rows as seven commemoration-reduced
and three transferred, counting 2038-06-23 in both halves to make the
arithmetic work. It is six and four; the row count itself was always
right, and is pinned by the comparator.
|
| | |
| |
| |
| |
| |
| | |
Agent research reports, SDD ledgers and review packages, same local-only
reasoning as docs/. Untracked but not ignored until now, so they showed up
in every git status.
|
| |/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
RG 33 omits a II/III-class vigil in three cases: it falls on a Sunday, it
falls on a I-class feast, "vel si festum cui praemittitur in alium diem
transferri aut ad commemorationem reduci contingat". Only the first two
were built; the third was recorded in precedence_ef.ml as unimplemented on
the grounds that no witness existed in the shipped data. That reasoning was
wrong, and the rule fires on 1 744 days across 1583-9999.
Both halves of the clause reduce to one observable question -- is the feast
the OBSERVED office on the following day (RG 34 puts it there) -- so the
kernel asks it once per candidate, after place_transfers has settled the
year. No fixed point is needed: a vigil is a candidate only on its own day,
never on its feast's, so suppressing it cannot change what the next day
observes. Precedence.rules gains vigil_feast, which returns the slug of the
feast a vigil precedes; the kernel cannot infer that itself, because only
two of the five vigil/feast pairs share a slug stem.
Blast radius, measured pre-change binary vs HEAD over the whole domain and
classified: 1 744 days, three shapes, zero unexplained. 1 199 are the feast
reduced to a commemoration (10 August on a Sunday, St Lawrence); 478 and 67
are the feast transferred under RG 96 after the Sacred Heart or Corpus
Christi takes its day. The Assumption's and the Ascension's vigils never
qualify -- their I-class feasts always keep their own day.
Independently witnessed, which is unusual here. The published Ordo -- the
only witness outside the Divinum Officium -> missalemeum -> lectio lineage
-- omits St Lawrence's vigil on 2025-08-09, agreeing with colitur against
both engines. That date had been read earlier as an Ordo gap; the Ordo was
right, and correcting the misreading is what surfaced this clause. On
2027-08-09 the feast does keep its day and the Ordo omits a vigil colitur
correctly keeps, which is a genuine Ordo gap.
Allow-lists: C39 (lectio, 10 rows) and a 2038 oracle class citing the
register, the 2026-2027 window having no instance. The golden pin asserting
St Lawrence's vigil is violet moved 2025 -> 2027; its own comment had
reasoned about the vigil's weekday and missed that RG 33 also looks at the
feast's. Two new pins cover both shapes of the clause.
The vigil/feast table is built with Slug.of_string_exn: mutation testing
showed that of_string plus Result.to_option turns a typo into None, which
this hook's contract reads as "not a vigil", switching the rule off in
silence. Two tests assert the table against the shipped data in both
directions.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
colitur.1 documented the flag and colitur-config(5) documented every
[defaults] key except this one, which the previous commit added.
Says what a flavour IS rather than only listing the values: the set of
characters that must be neutered so a value cannot corrupt its target.
A semicolon is ordinary text in LaTeX and a field separator in
iCalendar; an unescaped & or _ does not look wrong in LaTeX, it aborts
the build. Also why unset means INFER rather than 'no escaping', and why
there is no silent fallback.
|
| | |
|
| |
|
|
|
|
|
|
|
|
|
| |
The escaping flavour was the one rendering setting with no config key,
so a template whose extension says nothing needed the flag every run.
Unset still means infer, which config --show reports as (infer) rather
than (none).
Backfills the 0.8.0 and 0.9.0 changelog entries. Both were tagged by
hand rather than through 'make release', which is precisely the guard
that would have refused a release with no changelog line.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
The INI overlay form is a SOURCE format that 'colitur convert' turns into
the S-expression one. Handing the INI to --overlay instead failed deep in
the sexp reader with 'more than one S-expression in file', which names
neither the cause nor the cure -- and the shipped examples under
data/ef/examples/ are exactly the files someone would try that with.
The detector reads past the comment header to the first real line and
looks for a [section]. Its first version capped that scan at 40 lines and
poland.ini's own header is 43, so it silently never fired on the very
file it was written for; the cap now bounds only how much of a binary
gets read, and is not a guess at header length.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Three claims were false:
The download paragraph said older tags are not offered for download.
They are again -- per-tag snapshots were re-enabled, and only the
summary page's tag list is trimmed. A stable colitur-latest.tar.gz still
exists for scripts that should not know a version.
'Starting with the Roman EF (1962) and OF forms' read as though both
existed. lib/rites/ holds rite_ef and nothing else; the OF is the next
module. The tagline claimed multiple rites for the same reason.
Suite timings were from a much smaller suite: the fast run is ~8s over
568 tests, not ~3s, and the exhaustive sweep ~80s.
Adds a Status section with the real numbers, the five validation layers
and -- more usefully -- what they cannot see, since the differential and
the oracle share one data lineage. Notes that chants are deliberately
absent, and points at the colitur-x viewer.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
release-tarball chained its steps with ';', so a failed scp or ssh still
reached the final echo and it announced a publication that had not
happened -- the file was never in the docroot. Every step is now chained
with && and the result is checked over HTTP before anything is claimed:
the URL must respond and the bytes must untar.
Also passes -o ClearAllForwardings=yes, since the i2p port forwards in
the ssh config fail when another session holds them and their noise
buries real errors.
|
| |
|
|
|
|
|
| |
Requiring HEAD to be exactly a tag made the target unusable: the commit
that adds the download link to the README necessarily comes after the tag
it links to. It now archives the most recent tag and says so when HEAD is
ahead of it.
|
| |
|
|
|
|
|
|
|
|
| |
cgit offers snapshots for every tag or for none, so colitur.git now has
cgit.snapshots=0 and 'make release-tarball' uploads the tagged archive to
a stable URL. Ten tags had meant twenty download links, nineteen of them
stale.
The target refuses to run from an untagged or dirty tree, so what is
published always matches a tag.
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Robustness and hackability pass over the shipped program.
The citation parser is strict about numbers, four pairs of books no
longer share a full title, shipped styles re-parse their own output, and
overlay errors are written for the person editing the file rather than
for a compiler author.
colitur-templates(5) already refers to "before colitur 0.9" for the
month-key change, so the version moves with it.
|
| |\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
An audit of the shipped program, plus the fixes it found.
The citation parser accepted OCaml integer-literal syntax, so a typo like
'Luke 1_1:5' silently became a different chapter. Four pairs of different
books shared a full title -- 1 and 2 Corinthians both rendered 'Epistola
ad Corinthios' -- leaving 108 citations in 2027 alone that a reader could
not resolve to a book. Spec section 8.5 is now delivered rather than
recorded: shipped styles re-parse their own output. Overlay errors no
longer name OCaml source files at the reader.
Also: the new-overlay scaffold shows citations at the right nesting
level, config --show validates before printing, error messages no longer
echo whole file lines, and a month answers to both spellings of its own
name.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
A week object carries month_name/month_num/month_abbr, because at that
level a bare `name` would be ambiguous. An author who learned those names
inside {{#weeks}} reaches for them one level up inside {{#months}} too --
where they resolved to nothing, and an unknown key renders as the empty
string by design, so the result was a silently blank month heading rather
than any error.
Found by writing a template from scratch rather than copying a shipped
one; every shipped template sidesteps it, which is why nothing caught it.
The aliases are additive, so no existing template or golden changes.
colitur-templates(5) now documents both spellings and says why they
exist.
|
| |/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Found by auditing the shipped program rather than the diff.
The parser accepted OCaml integer-literal syntax, so "Luke 1_1:5" read as
chapter ELEVEN and "+5" as 5 -- a typo silently becoming a different
chapter, reachable through any user overlay. Numbers are now plain digits
and positive, and a descending range is rejected: 1:20-10 is always a
transcription error. No shipped citation changed.
FOUR PAIRS OF DIFFERENT BOOKS SHARED A FULL TITLE. 1 and 2 Corinthians
both rendered "Epistola ad Corinthios", as did Thessalonians, Timothy and
Peter -- 108 citations in 2027 alone that a reader cannot resolve to a
book. This is the Kings defect fixed earlier and not generalised. The
titles now carry their volume numeral, marked CONSTRUCTED, and a test
asserts no two books share a name -- while allowing the case where two
ids ARE the same book under different numbering, which a tradition
relates.
Spec section 8.5 is now delivered rather than merely recorded. Shipped
styles did not re-parse their own output: 32 of 52 Latin abbreviations
and 49 of 52 full titles failed, so a citation copied from colitur's own
output into an overlay was passed through untouched and printed in the
wrong language, silently. Every shipped name is registered as a spelling
and split_book learned multi-word titles by longest-token match. Now 0
of 52 fail beyond the same-book aliases.
Overlay errors were written for a compiler author: they named an OCaml
source file the reader does not have and buried the useful token. The
existing five-path rewriter is replaced by a generic one, applied to
every load path rather than one, so "rank: is not one of the allowed
values (at Class9)" replaces the raw Of_sexp_error dump.
Also: the new-overlay scaffold documented citations and layer without
showing them, and its comment implied the wrong nesting -- the single
easiest thing to get wrong; error messages echoed whole file lines,
copying an unrelated file's contents into stderr when a flag pointed at
one; and config --show validated partway down its table, exiting 2 after
writing five rows to stdout.
|
| |
|
|
|
|
|
|
|
|
| |
Citations become configurable: parsed into structure and re-rendered, so
book names, abbreviations, punctuation style and numbering tradition are
files a user edits rather than strings frozen in the data.
New lib/citation (Book, Parse, Render, Sigla); [bible] and [sigla]
sections in language files; lang/traditions.ini; three config keys and
CLI flags. --raw emits every citation byte-for-byte as stored.
|
| |\
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Citations are parsed into structure and re-rendered, so book names,
abbreviations, punctuation style and numbering tradition become files a
user edits rather than strings frozen in the data.
New lib/citation (Book, Parse, Render, Sigla); [bible] and [sigla]
sections in language files; lang/traditions.ini for numbering; three
config keys and CLI flags. --raw emits every citation byte-for-byte as
stored, bypassing the whole pipeline, so output stays diffable against
lectio and the raw view does not depend on the parser being correct.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The branch was RED and reported green. `dune test` exited 1: test/cli.t
pinned the pre-fix output `kings_1 19:3-8`, which the previous commit had
already fixed to `1 Reg 19:3-8`. The gate command piped dune through
`tail`, so it reported tail's exit status, and cram prints its diff
BEFORE the alcotest summary, so the two lines shown were the passing
ones. Verify with `dune test; echo $?`, never through a pipe.
A style file's own `book` key was unreachable. sigla_book resolved
against a hardcoded "abbr" and the result was applied unconditionally,
so the documented `[sigla] book = full` could never win. Render gains
book_string, and the style's own value is now the default that a flag or
config overrides. The unit test pinned style_of_fields correctly while
the wiring defeated it.
`lang --check` filtered the reference set to the celebration prefix, so
a file with no [bible] section at all reported a clean bill of health --
contradicting both the reason the keys change was made and lang.ml's own
comment. It now reports missing book names too.
The token test missed a FOURTH citation-bearing file: adjustments.sexp
writes citations as `Set_citation`, not `(reference ...)`. Its 16
citations all parse, so nothing was broken, but nothing was checking.
The first attempt at this fix read the file and extracted NOTHING -- the
marker stopped before the opening quote, so every payload was the part
label -- which is recorded in the code rather than left as a trap.
Also: colitur-config(5) claimed a trailing period the data does not
carry, and two la.ini scan quotes silently corrected OCR damage
("Ionae 3, I - I O", "Epistolse") while presenting themselves as
verbatim. Both are now marked as corrections.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Three defects, all found by running the documented examples rather than
by reading code.
All seven modern-numbering targets carried their own internal id as
their Latin name (kings_1.full = kings_1), so --sigla-tradition modern
printed "kings_1 19:3-8" -- leaking a key book.mli states is never shown
to a reader. They now carry Vulgate-derived Latin names marked
CONSTRUCTED: the 1962 Missal uses Vulgate numbering throughout, so it
can contain no incipit for a book that exists only under a later
convention. In Latin the modern tradition therefore only moves Kings and
Esdras; Osee, Ionas, Ecclesiasticus and the Apocalypse keep their names
either way, since modern numbering is a vernacular convention.
The coverage test could not catch this. It compares a name against its
KEY ("kings_1.full"), and "kings_1" differs from that, so the row
passed. A second check now asserts no name IS its own id. The two are
complementary: one catches a missing row, the other a row present but
filled with the wrong thing. Mutation-proved.
--help still claimed sigla settings were resolved "but not yet
rendered", stale since the render wiring landed. colitur-config.5
documented none of the three settings even though config --show prints
them.
|
| | |
| |
| |
| |
| | |
Includes that --raw emits citations verbatim, which is the behaviour
someone diffing against lectio depends on.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Both kings_3 and kings_4 carried the full title "Liber Regum", so
--sigla-book full rendered 3 Kings and 4 Kings identically and a reader
could not tell which book a citation named. Seven citations in 2027
alone.
The cause is real and now recorded rather than papered over: this
Missal's incipit "Lectio libri Regum." is UNDIFFERENTIATED -- that exact
unqualified phrase precedes both a "3 Reg." and a "4 Reg." locator in
the same book -- so it is not a valid full title for either volume on
its own. Only the numbered locator distinguishes them, and both "3 Reg"
and "4 Reg" are independently attested in the scans.
The full titles are therefore marked CONSTRUCTED, a third category
beside sourced and UNSOURCED: both halves are transcribed, their
combination is not. The numeral is written in Roman to match this
project's own output convention ("Feria IV", "Hebdomada I").
Abbreviated and --raw output are unchanged.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
la.ini and en.ini both gain [sigla] (the current Vulgate/Latin punctuation
convention, byte-identical to Render.default_style) and [bible] (a .full
and .abbr row for every id in Book.all -- 45 cited ids plus the 7 tradition
targets, 52 total). Shipping [sigla] changes no output, asserted by the
full test run. Shipping [bible] does change rendered book names, which is
the point.
la.ini's titles are sourced from docs/research/scan1.txt/scan2.txt (the
1962 Missal scans), each row citing the line its incipit pattern was read
from. Six pairs (kings_3/4, corinthians_1/2, thessalonians_1/2,
timothy_1/2, peter_1/2) share one Missal incipit and differ only in the
sourced .abbr, matching what the primary text itself does. Three ids
(proverbs, song_of_songs, ecclesiasticus) and the seven tradition targets
are marked UNSOURCED and fall back to the data's own spelling, per this
project's central rule against inventing a Latin title.
Re-running the task's own sourcing note against the scans, counting every
hit rather than eyeballing a frequency-sorted list, found nine of its
NOT-sourced verdicts were undercounted (a single clean hit, buried under
higher-frequency matches): Galatians, Colossians, both Thessalonians, both
Peter, Malachi, Numbers, Jonas, Osee and Esdras all have a clean incipit
in the scans and are sourced here. The note's other three verdicts stand,
confirmed independently. Ecclesiasticus is not simply unfound: this Missal
reuses Wisdom's own Lectio libri Sapientiae incipit for Ecclesiasticus
readings too (both were anciently classed as one Sapiential group), so
using it for Ecclesiasticus would misidentify the book, not merely
abbreviate it -- recorded in that row's own comment.
en.ini's [bible] is filled in full, not left partial the way [celebration]
is: without it, a Vulgate-numbered id would fall through the [meta]
fallback chain to la.ini's Latin name, not merely a less complete English
one. Traditional Douay-Rheims names for the Vulgate ids (3 Kings, Osee,
Ecclesiasticus, Isaias, Apocalypse), modern names for the seven tradition
targets, since --sigla-tradition modern is the reader asking for modern
numbering.
test_lang_coverage.ml gains test_every_book_named, asserting every
Book.all id has both forms in la.ini -- the check that catches a forgotten
tradition target, since nothing else in the suite ever names them.
Shipping real book names changes 11 golden template renders (Render/golden)
and several test/cli.t examples that used to demonstrate the pre-Task-10
default-spelling fallback; both are updated to the new, correct output,
each line checked against a fresh render before promoting.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
View.citation_ref and readings' part_ref are the only two places a
citation reaches output; every emitter goes through View.
--raw passes Sigla.verbatim rather than a style built over Lang.raw:
an identity name table would still reformat punctuation and renumber.
Loosen reject_sigla_for: --sigla-style/--sigla-book/--sigla-tradition
now actually render on readings/table/render/emit/publish (each
builds its own Sigla.t via the new load_sigla), and stay refused only
on commands that render no citation (day, easter, temporal, ...).
names_of degrades a Lang.bible miss to Book.default_spelling rather
than printing the lookup key itself ("luke.abbr"); pinned in cli.t
against a language file with no [bible] section at all, independent
of la.ini/en.ini's own eventual [bible] section.
Regenerate the 11 golden templates for 2027: every changed line is
one canonical citation replacing a stored variant, verified against
the full readings 2027 diff (raw vs default) cell by cell -- 86
distinct (raw, rendered) pairs account for the entire diff across all
11 files, with zero unexplained residue. Two families: the seven
duplicate book spellings collapsing onto one canonical form (Isa./Isa,
3 Kgs./3 Kings, ...), and citations reconstructed from parsed
structure dropping stray punctuation the parser already treats as
noise (a trailing period or semicolon, a comma chapter/verse
separator, an elided inherited chapter) -- each of the latter already
named and tested in test_citation.ml's parse_suite before this task.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
Same flag > config > default precedence as --lang, and each reported by
config --show with its source, so an override is visible rather than
mysterious.
A language file's [sigla] section IS a style; a config key SELECTS one and
may override settings within it -- the two are not a duplicate setting.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Apocalypse -> Revelation was missing from [modern] while `revelation` was
already a declared tradition target, so the target existed with nothing
mapping onto it. Same Vulgate-to-modern shape as Ecclesiasticus -> Sirach.
The injectivity test used to hand-copy the mapping table, which asserted
things about the copy and nothing about the artifact: a mapping added to
traditions.ini and forgotten in the test would have passed. It now reads
lang/traditions.ini and additionally asserts that [vulgate] is empty --
an entry there would silently renumber the DEFAULT, the one thing this
design promises never happens -- and that no declared target is left
unreachable.
Mutation-proved three ways: declaring an unused target, deleting a
mapping, and adding an entry to [vulgate] each redden it, the first two
naming the offending book.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Which book a reference denotes does not vary by language, so it lives in
its own file rather than in la.ini beside [weekday].
Six mappings, not four: Osee/Jonas are the same question as 3 Kings in
transliteration form.
The install rule is added in the same commit deliberately -- lang/ once
shipped without one, so the feature worked from the source tree and was
broken once installed.
bin/dune gains colitur_citation as a linked library, needed for the new
loader; lang_list's directory scan now excludes traditions.ini, which is
not a language file and does not parse as one.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Lang reads a hardcoded list of section names and ignores anything else
silently, so without this a [bible] section would appear to work and do
nothing -- confirmed directly with a scratch executable: before this
change, of_string on text containing a [bible] section parsed without
error and Lang.keys came back empty, no trace of the section anywhere.
Lang.bible keeps the total-lookup contract every other lookup in this
module has: a miss returns the key itself, never the empty string.
Lang.sigla_fields returns the [sigla] section's raw fields, values
trimmed but still quoted, for Render.style_of_fields to unquote.
[bible] joins the keys reference set so lang --check reports missing
book names; [sigla] deliberately does not, being settings with working
defaults rather than translatable names -- adding it would make --check
demand five settings from every language file.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
format never raises and returns an unparseable citation unchanged, so a
gap degrades to today's behaviour rather than to a crash. The coverage
test asserts separately that no shipped citation takes that path.
verbatim is what --raw uses: an identity name table would still reformat
punctuation and renumber, which would break byte-exact diffing against
lectio.
The existing citation-coverage walk (1970-2070) now also drives a
round-trip check in the same pass: parse -> render -> parse must reach
the same structure, using Book.default_spelling for names (Book.to_string
returns the internal id, which is not a registered token and cannot be
read back). Proved with a mutation: changing the default style's
part_sep to a separator the parser does not accept reddened the
round-trip case on 31 multi-part citations and nothing else; reverted.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Book.default_spelling returns the first registered spelling for an id. It
is the fallback display name, and it exists because the alternative is
worse: a language file's [bible] lookup is total and returns THE KEY on a
miss, so a book with no entry would render as "luke.abbr 5:12-14".
Falling back to the data's own spelling makes it render as "Luke
5:12-14" instead -- what colitur printed before this feature existed.
The degraded case is the old behaviour, the same principle Lang states
for its own key-returning misses.
Two test gaps closed, both found by mutation rather than by reading:
Parse's split_book scans a leading ordinal digit over '1'..'4', and no
case in the suite used an ordinal above 1. Narrowing the range to
'1'..'3' passed every test while seven real citations depend on it
("3 Kings 17:8-16", "4 Kings 5:1-15"). Parse-layer cases added; the
first attempt at this test asserted through Book.of_token, which is a
table lookup and never reaches split_book at all.
default_spelling is asserted to round-trip: every cited id's fallback
spelling must itself resolve back to that id, or Render and Parse
disagree the moment a book goes unnamed.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
A style is a set of format strings, so punctuation convention is data.
Values are unquoted here rather than in Overlay_ini: that parser trims
every value and is shared with overlays and [defaults], so teaching it
about quotes would change behaviour this feature has no business
changing.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
Walks 1970-2070 and parses every citation the engine emits. Nothing
validated citation spellings before this, which is why seven books
carried two spellings unnoticed.
Mutation-proved: removing the 'Isa.' spelling from the book table
reddens this test naming that citation.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
The parsed form is a book and a LIST of chapter-parts: the data cites
across chapters and lists disjoint verse ranges within one.
Two rules the shipped data forces and that are not obvious: a
semicolon-separated part may inherit the previous chapter rather than
restate it, and a chapter may be separated from its verses by a comma.
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
The book table was built against data/ef/lectionary.sexp alone. That
undercounts: sanctoral.sexp carries more citations than the lectionary
and, together with commons.sexp, cites 21 book tokens the table was
missing entirely (62 distinct tokens across all three files, not 42),
several of them common (2 Tim appears 8+ times).
Add the missing spellings to existing ids (2 Cor, Col., Exod, Ezek,
James, Sir, Eccli) and ten new ids for books not cited before (1-2
Timothy, 2 Peter, Apocalypse, Judith, Malachi, Proverbs, Song of Songs,
Tobit, Wisdom). Sir and Rev are modern spellings sitting inside
Vulgate data, so both resolve to their Vulgate ids (ecclesiasticus,
apocalypse) rather than to the sirach/revelation tradition targets --
mapping them to a second id would double-map the same book.
Add a duplicate-spelling invariant test (List.assoc_opt would silently
prefer the first match on a collision) and a test that re-derives the
token set from all three data files at test time and asserts every
token resolves, rather than trusting a survey performed once by hand.
|
| |/
|
|
|
|
|
|
|
| |
Seven books arrive in two spellings, inherited from lectio's ini and
ultimately from Divinum Officium. Collapse them onto one id here rather
than editing generated data.
Naming and renumbering are kept apart: a tradition decides which book an
id denotes, a language file decides what it is called.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ordo.ms's own Build: comment omitted -P-pa5 entirely, so the documented
command produced A5-sized content on a physical A4 PDF page
(595x842pt instead of 419x595pt) -- verified by running the command
verbatim and reading pdfinfo. Both .ms Build: lines were also missing
-Kutf8 and -t, which the Makefile's check-templates target already
carries: without -Kutf8, accented Latin names (e.g. Peñafort) render
as mojibake, confirmed by extracting text from the rendered PDF.
Bring both templates/ef/ordo.ms and templates/ef/grid.ms in line with
the Makefile's actual invocation (-ms -t -Tpdf -Kutf8 <papersize>), so
a user copying the Build: line gets the same result as
'make check-templates'. Updated the two golden fixtures that pin these
templates' rendered output byte-for-byte to match.
Verified every documented build command (README's two pdflatex
invocations, both .ms lines, both .typ lines) by running it verbatim
and checking pdfinfo/pdftotext output; no other flavour's Build: line
or man page example was affected.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
ordo.tex was 53 pages / 6.9 days-per-page on 2027 against ordo.typ's 27
pages / 13.5 -- the user prefers the tighter Typst booklet. Read ordo.typ's
own typographic decisions (base size, the day-number/name/detail-line size
hierarchy, box padding, inter-box skip, leading) and matched them: an
explicit 9pt/10.5 base (was \normalsize, 10pt), the day number and feast
name both 9pt bold (bold distinguishes them, not size, as ordo.typ decided),
7pt weekday, 6.5pt detail lines, sharp-cornered boxes, and needspace amounts
re-measured for the shorter boxes. Result: 29 pages / 12.6 days-per-page --
within the "a page or two of 27" target, reached without sacrificing
legibility (verified against the rendered PDF, not assumed).
The one real gotcha: \fontsize set inside a brace group that closes before
the line-ending \\ has no effect on the interline glue TeX inserts for that
break -- the glue uses the OUTER scope's \baselineskip, not the just-closed
group's. This is exactly the shape of bug the pre-retune box already had
(every \fontsize group closed before its own \par), which is why the first
retune pass barely moved the page count. Fixed by declaring \fontsize/
\selectfont/\bfseries/\mdseries inline, ungrouped, immediately before the
text each governs. Recorded in the file's own header for next time.
ordo.ms looked comparatively fine (43 pages) but was rendering on A4, not
A5 -- roughly double an A5 page's area, so normalised for area it was
already looser than ordo.tex's own pre-retune baseline. Fixed, each found
by rendering and reading the real PDF rather than assumed: added -P-pa5 to
the Makefile's groff invocation (mirroring grid.ms's existing -P-pa4l);
fixed the day box reading the RAW rank/colour strings (class-2, green)
instead of the resolved Latin pair (rank_name/colour_name) -- exactly the
"printed slugs" defect the ordo rebuild's own commit message named, which
this file had regressed to; added -Kutf8 (groff's default input encoding
produced visible mojibake on an accented Latin name); tightened fonts/
leading the same way as ordo.tex; found and fixed a stray-blank-line bug
present since the file's original authoring (every day/week/month boundary
left an accidental blank template line, read by groff as its own extra
.sp); found and fixed a genuine print-safety bug (text sitting 0.13mm from
the true page edge -- .pl, groff's text page length, is independent of
-P-pa5's device paper selection and needed matching explicitly before ms's
own FM-based bottom margin would fire at all); and added a .ne guard so a
week header can no longer be stranded alone at a page's foot. Net: 43 pages
(A4, broken) to 39 pages (real A5, print-safe, real Latin names, correct
encoding, zero orphans) -- no numeric target set for this flavour, but no
lower number was reachable without leaving one of the above unfixed.
ordo.html and every grid.* flavour are untouched, as scoped.
Both golden/ordo-2027.{tex,ms} regenerated through the real test harness
path (Test_support's loaders + Test_view's en-with-la-fallback view_of),
not the CLI, whose own --lang default (la) differs from the harness's and
would have pinned content the suite never produces.
Full suite green: 503 tests (504 with the exhaustive sweep). make
check-templates/check-schema/check-citations all clean.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Every ordo booklet week header used to print the month name again
even though the month heading right above it already established it
-- pure noise, repeated once per week for the whole year. All seven
ordo flavours now print the week's own Roman numeral plus the span of
dates it covers instead, e.g. "Hebdomada I (Ian 1-2)", with no month
name on the line; the LaTeX table of contents and the Typst outline
(built automatically from its own headings, so editing the heading is
the whole fix there) read the same way.
A single-day week (e.g. a lone trailing Sunday) renders "Ian 31", not
"Ian 31-31": each flavour's dash sits inside an inverted single_day
section, so the engine -- which cannot itself compare first_dom to
last_dom -- never has to decide anything, it only follows the data.
The dash character matches each flavour's own existing convention:
LaTeX/Typst's own "--" ligature, groff's \(en escape (alongside its
existing \(bu), HTML's – entity (alongside its existing
·), and a literal en dash for the three flavours with no
escaping (adoc/md/txt).
Every in-template comment describing the new fields is written without
ever typing two curly braces together, including in the two comments
this touches that used to embed a live {{month_name}}/{{term.week}}
tag pair inside a LaTeX % comment -- harmless only because that whole
line already started with %, not because the engine has any comment
awareness (it has none).
Goldens regenerated through test_render_golden.ml's own render path
(Test_view.view_of + Template.render_string), not the CLI -- the two
differ in default language and would otherwise pin output the test
suite never produces. Grid goldens are untouched, matching the grid
templates, which have no week header.
Verified against real pdflatex (two passes) and typst compile output,
not merely the golden pins: both flavours print "Hebdomada I (Ian
1-2)", "Hebdomada II (Ian 3-9)", "Hebdomada III (Ian 10-16)" as their
first three week headers and identical text in the table of
contents/outline, and both show "Hebdomada VI (Ian 31)" for January's
own trailing single-day week with no dash.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The ordo booklet's week header repeated the month name on every week
even though the heading above already established it -- pure noise.
Replace it with a Roman week number plus the span of dates the week
covers, e.g. "Hebdomada I (Ian 1-2)", following the project's own rule
that a presentation choice is data, not code.
lang/{la,en}.ini gain a [month_abbr] section (three-letter month
abbreviations); Lang.month_abbr follows Lang.month's exact shape,
including the out-of-range and miss-returns-the-key contracts. The
coverage test now fails loudly if an abbreviation goes missing, the
same as [month] already does.
Every week object in the view gains num_roman (Roman numeral, num
stays as the arabic original -- Roman is a presentation choice, not an
engine change), first_dom/last_dom (the day-of-month of the week's
first and last IN-MONTH days, padding excluded), month_abbr (resolved
through Lang.month_abbr), and single_day (true when the week holds
exactly one in-month day). single_day is a flag, not a preformatted
span string: the engine is logic-less and cannot itself decide between
"Ian 1" and "Ian 1-2", so a template makes that call from the flag
instead -- the same "shape the data, not the template" discipline
in_month and last already follow.
Weeks are built per month with padding only at the two ends, so a
week's in-month days never cross a month boundary -- verified, not
assumed: every week always has at least one real day since no month is
shorter than a single week.
Covered by three new View tests, including a real single-day-week
witness (January 2027's own trailing week is a lone Sunday, the 31st).
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
grid.typ used rows: (auto, 1fr) on the assumption that a Typst
fractional row always shrinks to share whatever space is left. It
does not: an 'fr' track's size is the LARGER of its fair share and
its content's own minimum, so it grows PAST 1fr when content is
taller than that share rather than compressing to fit. Adding the
sigla line (Epistle/Gospel references) grew every cell's natural
content past what a 1/6-of-remaining-space row allows, and a 6-week
month (three of twelve in 2027) split its own table between rows,
stranding the last week alone on a fresh page with no heading -- the
user's screenshot showed a row sliced mid-cell, 'I classis' cut off,
the next month's own heading printed underneath it.
Fixed the same way grid.tex's own \cellh already was: an explicit
row height (cellh, computed from the page's text height divided by
six, budgeted with 6mm of deliberate slack after a zero-slack first
attempt still split a table between rows), with each cell's content
wrapped in a height-clipped block sized to match -- content that
still does not fit is truncated rather than spilling into the next
cell or page. block(breakable: false) around the whole table is a
second, independent backstop: even if the budget is ever wrong
again, the WHOLE table jumps to a fresh page rather than splitting a
row across two, a loud failure (an odd blank page) rather than a
silent one (a sliced row).
This reverts the file's own prior divergence from grid.tex (Typst
stretching every month to fill the page regardless of row count,
disclosed as deliberate at the time) -- overflow safety now matters
more than that cosmetic fill. A 5-week month still leaves a blank
sixth row, the same trade-off grid.tex's own header already accepts.
Verified against all twelve months of 2027 (three 6-week: January,
May, October; nine 5-week), typeset through typst compile: exactly
12 pages, one month title per page, no clipped or duplicated
content, day 31 of a 6-week January fully visible. grid.tex itself
re-checked for the same latent risk and found unaffected -- its
fixed-height \cellh box already had enough margin.
Golden regenerated through the same Test_view.view_of path the
other affected templates already use.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
One week per page filled roughly the top third of every A5 page and
cost 65 pages for a year that needed far fewer -- the original
requirement, reversed here: a full 7-day week already occupies most
of an A5 page even at reduced type, so the real saving was always in
the year's many SHORT boundary weeks (a month's first/last week is
rarely a full seven days), not in packing two full weeks together.
lib/render/view.ml gains a [first] field on every week object (true
on a month's own first week, the same 'cheap flag beats invented
template logic' call [last] already made for grid rows) -- there is
still no {{../}} parent-path syntax for a template to test this any
other way.
ordo.tex: the per-week \clearpage is gone. Every week keeps its
existing small header line; the month's own first week additionally
prints a large, ruled banner. \needspace (a template-local LaTeX
dependency, not one of the frozen OCaml deps) keeps a header from
being stranded alone at a page's bottom -- and a real, separate bug
surfaced building it: \label placed BEFORE \needspace captured the
OLD page number when \needspace went on to force a break, producing
a step-stale \pageref in the TOC (fixed by moving \label after both
needspace checks). The day box itself is tightened (rank/colour
folded onto the same line as the citations -- an A5 box is far wider
than tall, so this fits even Holy Week's longest citations with room
to spare) so a second week routinely shares a page with the first.
Measured result: 65 -> 53 pages, two pdflatex passes still settle the
TOC, zero orphaned headers, zero overfull boxes.
ordo.typ: the per-week #pagebreak is gone; #show heading's own
block gains sticky:true, Typst's built-in answer to what ordo.tex
needs \needspace for. The month banner needed its OWN sticky:true,
found live: without it a banner could sit alone at a page's bottom
with the entire month pushed to the next page -- not caught by a
check for an orphaned HEADING, since the banner is plain markup, not
one. Measured result: 65 -> 27 pages (Typst's own page-fill packs
tighter than ordo.tex's needspace-driven approach), one compile pass,
zero orphans.
The five non-paginated flavours (ms/html/md/adoc/txt) never had a
forced page break to remove -- they mainly needed the same header
treatment for structural consistency: a week-level header nested
under each month's existing one, walking months>weeks>days (the
week's own padded, in-month-guarded day list) in place of the old
months>days flat loop. Verified against groff, asciidoctor and an
HTML parser -- zero warnings.
Goldens regenerated through the Test_view.view_of path
test_render_golden.ml itself uses, not the CLI (whose --lang default
would have pinned output the tests never produce).
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The booklet's title page and running header read "Ordo 2027 . ef" --
an internal identifier reaching a reader, the same defect class the
sanctoral/temporal slugs were already fixed for.
lang/la.ini and lang/en.ini each gain a [rite] section mapping the
rite id to a reader-facing name, sourced from the Missal's own title
page (docs/research/LT.txt:6 "MISSALE ROMANUM", :15 "EDITIO TYPICA")
plus the 1962 dating this project uses throughout (rules-register.md,
CLAUDE.md) -- both cited in the ini comments, marked PATTERN since the
phrase combines two title-page elements rather than quoting one
verbatim heading.
Lang.t gains a rite table/lookup (same total, miss-echoes-the-key
contract as every other lookup here); View.of_days exposes rite_name
alongside the existing rite field, which stays the stable key exactly
as slug is kept beside name. Every shipped template that printed
{{rite}} now prints {{rite_name}} instead.
test_lang_coverage.ml gains a coverage assertion over every rite id
the engine can emit -- one today -- so a second rite module (OF)
landing without a matching [rite] entry fails loudly instead of
printing its own bare id.
Goldens regenerated through the Test_view.view_of path
test_render_golden.ml itself uses (English with a Latin fallback),
not the CLI (whose --lang default is plain Latin and would have
pinned output the tests never produce). test/cli.t's own JSON prefix
assertion updated to match: rite_name is a real new key in that
generic view dump, sitting right after rite.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The task this branch of work set out to test was whether the template
engine is genuinely flavour-agnostic outside lib/render/escape.ml. It
is not, quite: bin/main.ml's own --flavour error/help text and both
man pages hand-typed the flavour list and its count ("the six
flavours") as literal prose, so adding Typst as a seventh left three
places silently wrong (an incomplete flavour list in two error
messages, and "six" where the true count is now seven) until this
commit.
bin/main.ml's two error messages (unknown --flavour value, and cannot
infer a flavour from an extension) now build their flavour list by
walking Escape.all/to_string instead of a literal string, so the next
flavour added will not reintroduce the same drift; the static --help
body text and colitur(1)/colitur-templates(5) still name flavours and
their count by hand; there was no equivalent gain available for
prose formatted this specifically, so those were updated by hand
instead, along with colitur-templates(5)'s FLAVOURS section, which
gained typst's own subsection alongside latex/groff/html/xml/ics/none.
test/cli.t's two pinned error-text assertions are updated to match,
and gained a new case: .typ inferred from extension, and a real
shipped citation's hyphen surviving --flavour typst's own escaping
live, the same shape the 2035 Ss. Petri et Pauli case already proves
for latex's ampersand just above it.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Adds templates/ef/ordo.typ (A5, one week per page, each day framed with
a colour swatch, real names, every fixed string through term.*) and
templates/ef/grid.typ (A4 landscape, one month per page, 7 columns,
colour-filled cells, Epistle/Gospel sigla), matching what ordo.tex and
grid.tex already do.
Typst resolves its own table of contents inside a single compile: no
pageref/aux-file dance, unlike pdflatex's own two-pass need for
ordo.tex. ordo.typ's TOC is one flat heading per week (Month . Week N)
rather than ordo.tex's own month-once/weeks-nested layout, since the
view model has no first-of-month flag for a template to test and
inventing one would be new engine state for a cosmetic grouping.
grid.typ deliberately diverges from grid.tex's own partial-fill
behaviour for a 5-week month: Typst's row-size list repeats its own
last entry for every remaining row, so a two-entry rows spec fills the
page completely regardless of week count, where LaTeX's fixed cellh
leaves a 5-week month's bottom sixth blank on purpose. Both divergences
are documented in the templates' own headers.
Two bugs found and fixed while building these against the real typst
0.14.2 binary: two stray hash characters in grid.typ's header comment
that were not actually inside a line-comment (would have been parsed
as Typst code, not prose); and two lines merging into one wherever they
were separated only by a bare newline (still the same soft-wrapped
paragraph in Typst, unlike LaTeX's own explicit paragraph break) --
fixed with a trailing backslash to force a real line break without
adding inter-paragraph spacing, verified against a full-year render
(zero typst warnings across all 365 days of 2027, in both templates).
Goldens (test/golden/ordo-2027.typ, test/golden/grid-2027.typ) were
generated through the harness's own View.of_days/Template.render_string
path, via a throwaway generator, NOT through the CLI: the CLI's default
language table differs from the harness's own English-chained-to-Latin
table, confirmed live (diffing CLI output against the harness's own
render showed exactly the expected language divergence and nothing
else) -- the same trap this project has hit before.
make check-templates now typesets both templates with the typst binary,
using the same SKIPPED-and-exit-0 shape the pdflatex/groff blocks
already use when their own tool is absent, verified in both directions:
absent (PATH without typst) skips loudly and exits 0, and a
deliberately corrupted template fails the target with a non-zero exit
and typst's own error text.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Adds Escape.Typst: type, all, to_string ("typst"), of_string,
of_extension (.typ), and a per-character expand function, structured
identically to the existing latex escaper (one pass, no re-scan, so
double-escaping stays impossible by construction).
The metacharacter set was verified against the installed typst 0.14.2
binary, not assumed: a probe document escaping each of #, *, _, $, @,
<, >, `, \, ~ and - was compiled and pdftotext'd back to confirm the
literal character survives, and each was separately confirmed to do
something else when left bare (# opens code mode, */_ toggle
strong/emph, $ opens math, @ opens a reference -- a bare unresolved
@word is a hard compile error, not merely mangled output -- </> can
close around a bare word into label syntax that swallows it whole, `
opens raw, ~ is a non-breaking space, and a run of two or three '-'
becomes an en/em dash). All ten are backslash-escapable; none needed a
non-backslash workaround. '-' is escaped unconditionally rather than
only inside a detected run, since this escaper has no lookahead -- a
probe confirmed escaping every hyphen independently still typesets as
literal hyphens for a run of any length, so the single per-character
rule is sufficient.
test_escape.ml's new Typst cases were written first and shown to fail
against a stubbed identity apply before the real escaper landed, per
this project's own regression-test discipline. Comments in both files
mark which new tests are genuine regression tests (the per-character
escaping, including a real shipped citation and a synthetic
markdown-habit overlay name) versus characterisation (to_string/
of_string/of_extension are flat table lookups with no logic to have
been wrong).
|
