TimedSchedule

Identified appointment spans in source-label order. The window selects starts; duration may extend beyond it. IDs preserve the series and source. Start and end interpretation share one zone-work budget per call. new_with_endings supports source-specific duration or explicit-end overrides.

Example

Use this layer when each recurrence start should become an identified span. Coordinate measures POSIX microseconds; Calendar applies calendar components, resolves the end, then adds its coordinate tail. The window selects starts, not every appointment overlapping that window.

This example supplies a finite UTC rules snapshot. Applications using named zones can pass rules from the optional zone-data package instead. For a complete application with calendar-month durations and an extra booking, see equipment loans.

import time.TimedSchedule
import time.TimedRecurrence
import time.TimedOccurrence
import time.CalendarPattern
import time.GregorianDate
import time.Calendar.Date
import time.ClockTime
import time.LocalDateTime
import time.ZoneRules
import time.FixedOffset
import time.PosixSpan
import time.PosixBoundary
import time.PosixDelta

expect {
    date = GregorianDate.from_fields({ year: 1970, month: 1, day: 1 })?
    end_date = GregorianDate.from_fields({ year: 1970, month: 1, day: 4 })?
    clock = ClockTime.from_microseconds_since_midnight(0)?
    start = LocalDateTime.new(Calendar.Date.from_gregorian(date), clock)
    end = LocalDateTime.new(Calendar.Date.from_gregorian(end_date), clock)
    validity = PosixSpan.from_seconds(-1, 345600, RejectSubmicrosecond)?
    rules = ZoneRules.new_bounded(
        "UTC", "fixed", validity,
        FixedOffset.from_seconds(0), [],
        { minimum: 0, maximum: 0 },
    )?
    rule = TimedRecurrence.new({ date, clock }, {
        calendar: CalendarPattern.defaults(Daily),
        clocks: { hours: [], minutes: [], seconds: [] },
        termination: Count(2),
        by_set_pos: [],
    })?
    duration = Coordinate(PosixDelta.from_microseconds(3600000000))
    schedule = TimedSchedule.new(
        42.U64, rule, { start, end }, duration,
        { rules, occurrence: RequireUnique, gap: RejectGap },
    )?
    work = {
        max_steps: 100, max_buffered: 1,
        max_zone_segments: 10, max_zone_candidates: 1,
    }
    result = TimedSchedule.collect(schedule, {
        work, max_occurrences: 3,
    })
    match result {
        Err(_) => Bool.False
        Ok(batch) => match batch.status {
            Complete => batch.occurrences.len() == 2
            Limited(_) => Bool.False
        }
    }
}

When a batch returns Limited, process its partial output and resume the returned cursor with sufficient budgets. Increasing only the output cap does not fix a work or buffer limit.