sopcore
True Sums of Products
Version on this page:  0.4.0.0 
LTS Haskell 13.22:  0.4.0.0 
Stackage Nightly 20190522:  0.4.0.0 
Latest on Hackage:  0.5.0.0 
Module documentation for 0.4.0.0
Implementation of nary sums and nary products.
The module Data.SOP
is the main module of this library and contains
more detailed documentation.
The main use case of this package is to serve as the core of
genericssop
.
A detailed description of the ideas behind this library is provided by the paper:
Edsko de Vries and Andres Löh. True Sums of Products. Workshop on Generic Programming (WGP) 2014.
Changes
0.5.0.0 (20190509)

Add
ejections
that computes a product of functions that try to extract an element out of an nary sum. (See #91.) 
Change the definition of
SameShapeAs
to be nonrecursive and thereby improve compiler performance. (See #105.)
0.4.0.0 (20181020)

Split into
sopcore
andgenericssop
packages. 
Drop support for GHC < 8.0.2, bump
base
dependency to>= 4.9
and remove dependency ontransformers
. 
Simplify
All2 c
toAll (All c)
and simplifySListI xs
toAll Top xs
, and some implied refactoring. 
Add
Semigroup
andMonoid
instances for various datatypes.
0.3.2.0 (20180108)

Make TH
deriveGenericFunctions
work properly with parameterized types (note that the more widely usedderiveGeneric
was already working correctly). 
Make TH
deriveGeneric
work properly with empty types. 
Add
compare_NS
,ccompare_NS
,compare_SOP
, andccompare_SOP
to better support comparison of sum structures. 
Add
hctraverse_
andhctraverse'
as well as their unconstrained variants and a number of derived functions, to support effectful traversals.
0.3.1.0 (20170611)

Add
AllZip
,htrans
,hcoerce
,hfromI
,htoI
. These functions are for converting between related structures that do not have common signatures.The most common application of these functions seems to be the scenario where a datatype has components that are all wrapped in a common type constructor application, e.g. a datatype where every component is a
Maybe
. Then we can usehfromI
afterfrom
to turn the generically derivedSOP
ofI
s into anSOP
ofMaybe
s (and back). 
Add
IsProductType
,IsEnumType
,IsWrappedType
andIsNewtype
constraint synonyms capturing specific classes of datypes.
0.3.0.0 (20170429)

No longer compatible with GHC 7.6, due to the lack of support for typelevel literals.

Support typelevel metadata. This is provided by the
Generics.SOP.Type.Metadata
module. The two modulesGenerics.SOP.Metadata
andGenerics.SOP.Type.Metadata
export nearly the same names, so for backwards compatibility, we keep exportingGenerics.SOP.Metadata
directly fromGenerics.SOP
, whereasGenerics.SOP.Type.Metadata
is supposed to be imported explicitly (and qualified).Termlevel metadata is still available, but is now usually computed automatically from the typelevel metadata which contains the same information, using the function
demoteDatatypeInfo
. Termlevel metadata is unchanged from genericssop0.2, so in most cases, even if your code makes use of metadata, you should not need to change anything.If you use TH deriving, then both typelevel metadata and termlevel metadata is generated for you automatically, for all supported GHC versions.
If you use GGP deriving, then typelevel metadata is available if you use GHC 8.0 or newer. If you use GHC 7.x, then GHC.Generics supports only termlevel metadata, so we cannot translate that into typelevel metadata. In this combination, you cannot use code that relies on typelevel metadata, so you should either upgrade GHC or switch to THbased deriving.
0.2.5.0 (20170421)

GHC 8.2 compatibility.

Make
:.:
an instance ofApplicative
,Foldable
andTraversable
. 
Add functions
apInjs'_NP
andapInjs'_POP
. These are variants ofapInjs_NP
andapInjs'_POP
that return their result as an nary product, rather than collapsing it into a list. 
Add
hexpand
(andexpand_NS
andexpand_SOP
). These functions expand sums into products, given a default value to fill the other slots. 
Add utility functions such as
mapII
ormapIK
that lift functions into different combinations of identity and constant functors. 
Add
NFData
(and lifted variants) instances for basic functors, products and sums.
0.2.4.0 (20170202)

Add
hindex
(andindex_NS
andindex_SOP
). 
Add
hapInjs
as a generalization ofapInjs_NP
andapInjs_POP
. 
Make basic functors instances of lifted classes (such as
Eq1
etc).
0.2.3.0 (20161204)

Add various metadata getters

Add
hdicts
. 
Add catamorphisms and anamorphisms for
NP
andNS
. 
TH compatibility changes for GHC 8.1 (master).
0.2.2.0 (20160710)

Introduced
unZ
to destruct a unary sum. 
Add Haddock
@since
annotations for various functions.
0.2.1.0 (20160208)

Now includes a CHANGELOG.

Should now work with ghc8.0.1rc1 and rc2 (thanks to Oleg Grenrus).

Introduced
hd
andtl
to project out of a product, andProjection
andprojections
as duals ofInjection
andinjections
.
0.2.0.0 (20151023)

Now tested with ghc7.10

Introduced names
hmap
,hcmap
,hzipWith
,hczipWith
forhliftA
,hcliftA
,hliftA2
,hcliftA2
, respectively. Similarly for the specialized versions of these functions. 
The constraint transformers
All
andAll2
are now defined as type classes, not type families. As a consequence, the partial applicationsAll c
andAll2 c
are now possible. 
Because of the redefinition of
All
andAll2
, some special cases are no longer necessary. For example,cpure_POP
can now be implemented as a nested application ofpure_NP
. 
Because of the redefinition of
All
andAll2
, the functionshcliftA'
and variants (with prime!) are now deprecated. One can easily use the normal versions instead. For example, the definition ofhcliftA'
is now simplyhcliftA' p = hcliftA (allP p) where allP :: proxy c > Proxy (All c) allP _ = Proxy

Because
All
andAll2
are now type classes, they now have superclass constraints implying that the typelevel lists they are ranging over must have singletons.class (SListI xs, ...) => All c xs class (SListI xss, ...) => All2 c xss
Some type signatures can be simplified due to this.

The
SingI
typeclass andSing
datatypes are now deprecated. The replacements are calledSListI
andSList
. Thesing
method is now calledsList
. The difference is that the new versions reveal only the spine of the list, and contain no singleton representation for the elements anymore.For onedimensional typelevel lists, replace
SingI xs => ...
by
SListI xs => ...
For twodimensional typelevel lists, replace
SingI xss => ...
by
All SListI xss => ...
Because All itself implies
SListI xss
(see above), this constraint is equivalent to the oldSing xss
.The old names are provided for (limited) backward compatibility. They map to the new constructs. This will work in some, but not all scenarios.
The function
lengthSing
has also been renamed tolengthSList
for consistency, and the old name is deprecated. 
All
Proxy c
arguments have been replaced byproxy c
flexible arguments, so that other type constructors can be used as proxies. 
Classlevel composition (
Compose
), pairing (And
), and a trivial constraint (Top
) have been added. Typelevel map (Map
) has been removed. Occurrences such asAll c (Map f xs)
should now be replaced with
All (c `Compose` f) xs

There is a new module called
Generics.SOP.Dict
that contains functions for manipulating dictionaries explicitly. These can be used to prove theorems about nontrivial class constraints such as the ones that get built usingAll
andAll2
. Some such theorems are provided. 
There is a new TH function
deriveGenericFunctions
that derives the code of a datatype and conversion functions, but does not create a class instance. (Contributed by Oleg Grenrus.) 
There is a new TH function
deriveMetadataValue
that derives aDatatypeInfo
value for a datatype, but does not create an instance ofHasDatatypeInfo
. (Contributed by Oleg Grenrus.) 
There is a very simple example file. (Contributed by Oleg Grenrus.)

The function
hcollapse
forNS
now results in ana
rather than anI a
, matching the specialized versioncollapse_NS
. (Suggested by Roman Cheplyaka.)
0.1.1.2 (20150327)
 Updated version bounds for ghcprim (for ghc7.10).
0.1.1.1 (20150320)

Preparations for ghc7.10.

Documentation fix. (Contributed by Roman Cheplyaka.)
0.1.1 (20150106)

Documentation fixes.

Add superclass constraint (TODO).

Now derive tuple instance for tuples up to 30 components. (Contributed by Michael Orlitzky.)