Revision history for red-black-record

2.1.3.0

• Added injections_Record.

2.1.2.0

• Deprecated VariantInjection and injections_Variant. VariantInjections is superfluous, as Case can be used instead.
• Added runCase accessor for Case.

2.1.1.0

• Added NP-like functions for working on Records, like sequence_Record. They are memebers of the Maplike typeclass.
• Deprecated the -Subset functions from Data.RBR and created a new module Data.RBR.Subset with new versions. To avoid collisions, Data.RBR.Subset should be imported qualified.
• Deprecated a number of other functions that had Productlike / Sumlike constraints, added new functions with Maplike constraints.
• Added IsRecordType, IsVariantType.
• Added KeyValueConstraints, ValueConstraint.

2.1.0.0

• Made the type-level map poly-kinded in the values, as there wasn't a real reason to force them to the Type kind.
• Removed deprecated EmptyMap (use Empty instead).

2.0.4.0

• Compatibility with sop-core 0.5.0.0.
• Contravariant intance for Case newtype.

2.0.3.0

• Issue #7: FromVariant & ToVariant instances for sum types with branches with no arguments.

2.0.2.2

• Improved compilation times for type-level deletion.

2.0.0.0

• BREAKING CHANGES

  • The constructors for the type-level map are now hidden.
  • The name of the type-level map has changed from RBT to Map, to de-emphasize implementation details.

• Added the "Deletable" typeclass with de "delete" and "winnow" methods.

• Solved bugs with coloring/balancing, added new tests.

• Data.RBR.Internal is still exported, but it doesn't appear in the Haddocks. It appears that Haddock doesn't play well with reexported-modules sections in Cabal.

1.1.0.0

• Field and Branch type families to help speed up type-level computations.

  Apparently, having identical invocations of a "costly to compute" type family in a signature slowed things down.

1.0.0.2

• Improved compilation times for getters by refactoring KeyHelper.

1.0.0.0

• First version. Released on an unsuspecting world.