Changelog

NEXT

• #114: fix doc typo (@metayan)
• #111: docstrings of java-time.api contain literal \n (@devurandom

1.4.2

• #105: fix not-{before,after}? on unconverted values
• a consequence of fixing #104
• add java-time.api/= for equality of times (with conversions)
• add support to intermix unconverted values in comparison operations after first argument
• supported by {before,after}?, not-{before,after}?, j/=, and aliases of those ops
• examples:
  • (j/< (j/day-of-week :thursday) :saturday :sunday)
  • (j/<= (j/day-of-week :thursday) :thursday (j/day-of-week :saturday) :sunday)
  • (j/= (j/day-of-week :thursday) :thursday (j/day-of-week :thursday) :thursday)

1.4.1

• #104: fix transivitity of not-{before,after}? when called with intervals

1.4.0

• #98: fix not-before? and not-after? with one or more than two arguments
• new aliases
• java-time.api/+ aliases java-time.api/plus
• java-time.api/- aliases java-time.api/minus
• java-time.api/neg? aliases java-time.api/negative?
• java-time.api/< aliases java-time.api/before?
• java-time.api/> aliases java-time.api/after?
• java-time.api/<= aliases java-time.api/not-after?
• java-time.api/>= aliases java-time.api/not-before?

1.3.0

• remove :redef on protocol methods, they are never direct-linked: https://ask.clojure.org/index.php/10967/are-protocol-methods-guaranteed-to-not-be-directly-linked?show=10990#a10990
• #100: respect *clock* when only providing a zone id in constructors

1.2.0

• #95 work around CLJ-1796

1.1.0

Due to #91, the main java-time namespace has been deprecated. A new namespace java-time.api has been created

Note that this change is entirely optional—java-time and java-time.api will continue to be in sync and may coexist.

See the docstring for java-time for more details.

1.0.0

Released 2022-11-24.

New Features

• Previously, Intervals were only allowed to the left of Instants in {before,after}?. Now they can be freely intermixed.

Fixed

• #78(liquidz): Add missing chrono fields (v0.3.3 has breaking changes)
• #81(terop): Remove clj-tuple - no advantages over Clojure vector anymore
• #52: Fix before/after on Intervals
• #83(imrekoszo): Exclude clojure.core/abs

Enhancements

• add docstrings to all java-time fns
• support clj-kondo in java-time ns by adding :arglists to all vars

Internal

• Deprecate java-time.util/get-static-fields-of-type
• Remove java-time.potemkin.namespaces

0.3.3

New Features

• #71(brettrowberry): not-after? and not-before?
• #40(davidneu)/#61(puredanger): added/fixed deps.edn

Fixed

• #72(FieryCod): working with GraalVM
• #29(danieldroit): conversion graph construction edge case
• #60(robdaemon): locales can mess up predefined-formatters
• #51(jimpil): remove reflection during load - improves load time

Docs

• #27(emlin)
• #35(thobbs)
• #36(holyjak)
• #38(ProjectFrank)
• #39(sashary)
• #48(bpringe)
• #56,#57(green-coder)
• #63(vandr0iy)
• #54,#31,#24

0.3.2

New Features

• zone-id? predicate
• set-clock! - sets the mocked clock value to the supplied instant
• when-joda-time-loaded - macro which runs code when Joda-Time is on the classpath
• instant->sql-timestamp - produce a java.sql.Timestamp from an Instant-like object
• as support for two-field Time entities #21, courtesy Larry Jones

0.3.1

New Features

• clock? predicate
• mock-clock - returns a mocked instance of java.time.Clock.

0.3.0

Breaking changes

• to-sql-date converts anything convertible to LocalDate into a java.sql.Date. Previously to-sql-date returned a java.util.Date (bug).
• #10 to-sql-timestamp stopped accepting an instant and starting accepting local date time. I mistakenly assumed that java.sql.Timestamp/from(Instant) was deprecated.

New Features

Please see a new section within README - Legacy Date-Time Types

• java-date, sql-date, sql-timestamp, sql-time - functions which produce the java.util.* date-time objects.
• #5 automatic conversions:
  • java.sql.Date -> java.time.LocalDate
  • java.sql.Timestamp -> java.time.LocalDateTime
  • java.sql.Time -> java.time.LocalTime
• deprecated to-java-date/to-sql-date/to-sql-timestamp

0.2.2

Fixed

• Wrong primitive type annotation on to-millis-from-epoch, see Eastwood docs for the explanation.

0.2.1

Fixed

• #1: Reflection warnings in two-field time entity constructors
• #2: Ordered implementation for java.time.Instant

0.2.0

Breaking changes

• zoned-date-time doesn’t accept the zone id as the last argument
• offset-date-time/offset-time doesn’t accept offset id as the last argument

New features

• with-offset/with-offset-same-instant for offset manipulation
• with-zone/with-zone-same-instant for zone manipulation