Nine Sundays went unnamed in every jurisdiction. Three were missing from the
movable cycle outright: the Paralytic, the Samaritan Woman and the Blind Man,
so the fourth, fifth and sixth Sundays of Pascha reported only the saint of the
fixed date. Six more are reckoned from a fixed feast rather than from Pascha —
before and after the Exaltation, the Forefathers, before and after the
Nativity, before and after the Theophany — and were not modelled at all.
The three Pentecostarion Sundays rank with Thomas Sunday and the Myrrh-bearing
Women, so a vigil-rank saint still keeps the day and the Sunday is commemorated
on it. The six fixed-cycle ones never name a day: they are found in
ecclesiastical space, so the old calendar and the new get them on the right
morning, and they join the commemorations. Where a short window holds no Sunday
the Typikon says what to do and this follows it; where the books disagree, the
day goes unmarked rather than carrying an answer this calendar cannot defend.
`week` says where a day falls in the count from Pentecost, which is how the
books head the long ordinary stretch. It runs on across the new year and stops
at the Triodion, whose Sundays have names instead of numbers.
`--every-day` gives ics an event for every day, ferias included. The default
still thins the calendar, which is right for a subscription and wrong for a
reader who wants the saint of each morning; and the description now leads with
the commemorations rather than burying them under the church date.
Checked against 1.1.2 over every jurisdiction from 2015 to 2035: rank, feast
and title move on the three new Sundays and nowhere else. No fast, obligation,
binding or season changes on any day.
Every desktop shell that reads typikon today runs the binary and parses
its JSON. That costs about 1.5ms and is the right answer for a calendar
that changes once a day, but it is the wrong shape for anything with an
FFI and no business spawning a process on every redraw.
Dates cross as Julian Day Numbers, so no struct layout is part of the
contract. A day is opaque and owns its strings, built once when it is
reckoned: an accessor returns the same borrowed pointer every time, so a
statusline polling one in a loop allocates nothing. The accessors carry
every field of the JSON, so nothing on the other side needs a parser.
The Python bindings and this were resolving a jurisdiction by two copies
of the same logic, and the copy here leaked one per call. Both now share
jurisdictions::resolve_full, which interns the combinations instead.
The header is a promise made in a second file, and a second file drifts:
a test reads both and fails if either has a symbol the other does not,
and does the same for the LuaJIT cdef, where a stale declaration is a
segfault rather than a compile error.
The years this calendar answers for are now 1 to 9999 and enforced at every
constructor. Arithmetic saturates instead of wrapping, which in a release
build had put a date in antiquity and produced an ISO string that was not ten
characters long; a seven-day walk stops at the boundary rather than running
past it.
--days took any integer and aborted the process trying to allocate for it, as
did --indent. Both are bounded now.
--update-menaion parsed a day out of a fetched page with no bound on it and
wrote keys like 01-4294967295 that no lookup could ever match, then wrote the
file even when the parse had failed, in replace mode, which would have blanked
a year of commemorations in a calendar that had been working. It bounds the
day by the month and refuses to write when most of the year comes back empty.
today() shelled out to date(1) and fell back to UTC without a word when it was
absent, so a container reported the wrong day either side of midnight. It says
so now.
The UGCC loaded the Ruthenian menaion. That is the Metropolia of Pittsburgh's
calendar of saints, and lending it to Kyiv is the same error as lending the
law, in a release whose headline is that a church's law is never borrowed.
Also: has_obligation was dead API about to be frozen, and two ptr::eq
comparisons stood in for asking whether the discipline was overridden, which
is what they meant.
Both keep the same Typikon and read it a little differently, and until now
this library silently picked one reading and called it the Typikon. Two
differences are modelled, each one visible in the output and named in the
day's reason:
- The Protection of the Theotokos is a great feast in Slavic usage. Greek
practice ranks it lower, and the Church of Greece keeps it on 28 October
besides. This has teeth: rank is what lifts a Wednesday or Friday fast.
- Greek practice admits fish on any day of the Nativity Fast but Wednesday
and Friday until 17 December; Slavic practice keeps fish to Saturdays and
Sundays throughout.
Every jurisdiction carries the usage its books follow, and --tradition
overrides it for a parish that does not match its jurisdiction's default. The
difference in the calendar lives in a data overlay compiled in beside the
common menologion, which is where a local church's own commemorations would
go; the difference in the fast lives in the rule, because that is where it is.
Each rank also now carries the mark the books print beside it, circled cross,
cross and Jerusalem cross for great, vigil and polyeleos. Unicode has no glyph
for the polyeleos sign, a cross over a half-circle, so the Jerusalem cross
stands in as digital calendars generally settle on.
Python was never the house language here: the mw-* desktop surface it came
from is 39 bash scripts and was exactly one Python script, the one this was
extracted from. What it cost was distribution, a widget dependency that
needed pipx and a Python floor, and 24ms of interpreter startup before a line
of the program ran.
The port is verified rather than asserted. Both implementations were run over
199,446 day-records, 26 jurisdictions across 21 years, with every field of
every record required to match; the JSON, the iCalendar and 480 renderings of
the text output are identical too. examples/dump.rs is the harness.
What the type system actually buys, which is less than the usual claim and
more than nothing:
- Level is an ordered enum, strictest first, so reads as 'a is
stricter' and relaxed_to() is the one place a feast can lift a fast. No
rule can get the direction backwards.
- Date is a newtype over the Julian Day Number and JulianDate is a separate
type, so a civil date and an ecclesiastical one cannot be mixed up. That
confusion is the single most common way calendar code goes wrong.
- Octets is a newtype in the iCalendar folder, because the bug it replaced
counted characters against a limit RFC 5545 states in octets.
- Movables is a struct, not a map, so a mistyped key is a compile error
rather than a None at the far end of the program.
Of the six defects an adversarial review found in the Python, exactly one had
a shape a type system would have caught. The rest were logic, and the tests
came with them.
The menologion is compiled in with include_str!, so the binary answers on its
own: 930KB, no runtime, 1.2ms to start against Python's 36ms.
funkpower