# LGtk/ADT lenses

### From HaskellWiki

(add link) |
m (a remark) |
||

Line 102: | Line 102: | ||

Remarks: | Remarks: | ||

− | * If the top level constructor of the list is <hask>[]</hask>, the head and the tail of the list can still be edited; the change will only be visible through <hask>listLens</hask> when the constructor is changed back to <hask>(:)</hask>. This may seem to be odd, but for many applications this is the right behaviour. | + | * If the top level constructor of the list is <hask>[]</hask>, the head and the tail of the list can still be edited; the change will only be visible through <hask>listLens</hask> when the constructor is changed back to <hask>(:)</hask>. This may seem to be odd, but for many applications this is the right behaviour. See the interpretation subsection below. |

* For editing the tail of the tail of the list, we need an <hask>s' :: S</hask> such that <hask>s</hask> viewed through <hask>tailLens</hask> is the same as <hask>s'</hask> viewed through <hask>listLens</hask>. Explained on a figure: | * For editing the tail of the tail of the list, we need an <hask>s' :: S</hask> such that <hask>s</hask> viewed through <hask>tailLens</hask> is the same as <hask>s'</hask> viewed through <hask>listLens</hask>. Explained on a figure: | ||

Line 108: | Line 108: | ||

On the figure, edges are lenses and nodes are ''references''. One possible definition of references is described in [[LGtk/Semantics#References]]. How <hask>s'</hask> can be created and how <hask>s</hask> and <hask>s'</hask> can be kept in sync is a related but separate question. [[LGtk/Semantics#Dependent_reference_creation]] describes a possible solution. | On the figure, edges are lenses and nodes are ''references''. One possible definition of references is described in [[LGtk/Semantics#References]]. How <hask>s'</hask> can be created and how <hask>s</hask> and <hask>s'</hask> can be kept in sync is a related but separate question. [[LGtk/Semantics#Dependent_reference_creation]] describes a possible solution. | ||

− | |||

=== Example: ADT with repeated record fields === | === Example: ADT with repeated record fields === |

## Revision as of 23:13, 7 June 2013

## Contents |

## 1 Problem description

Lenses provide uniform and compositional way to view and edit data structures.

For example, one can view and edit pairs withq == setL fstLens (getL fstLens q) (setL sndLens (getL sndLens q) p)

Similarly, there is a toolbox of lenses for records which toolbox contains one lens for each record field.

**Question:**

Are there a toolbox of lenses for algebraic data types with multiple constructors?

## 2 Existing solutions

### 2.1 Partial lenses

The data-lens library provides partial lenses which are isomorphic to

type PartialLens a b = (a -> Maybe b, a -> Maybe (b -> a))

The flollowing partial lenses are defined for lists:

headLens :: PartialLens [a] a headLens = (get, set) where get [] = Nothing get (h:t) = Just h set [] = Nothing set (h:t) = Just (:t)

tailLens :: PartialLens [a] [a] tailLens = (get, set) where get [] = Nothing get (h:t) = Just t set [] = Nothing set (h:t) = Just (h:)

### 2.2 Other solutions

*Please help to extend the list of known solutions.*

## 3 ADT lenses

The proposed solution, summarized:

**As a lens toolbox for an ADT, use a lens whose codomain is the ADT and whose domain is a tuple of the constructor tag and the ADT components.**

Let's see specific examples before the generic descripton of the proposed lens.

### 3.1 Example: List lens

The lens for lists which forms a complete toolbox:

`import Data.Lens.Common`

listLens :: Lens (Bool, (a, [a])) [a] listLens = lens get set where get (False, _) = [] get (True, (l, r)) = l: r set [] (_, x) = (False, x) set (l: r) _ = (True, (l, r))

#### 3.1.1 List lens usage

Suppose that we have a statetype S = (Bool, (Int, [Int]))

We can view and edit the list through the following lenses:

- edits thelistLens :: Lens S [Int]
**complete list**. - edits thefstLens :: Lens S Bool
**top level constructor**of the list:corresponds toFalseand[]corresponds toTrue.(:) - edits theheadLens = fstLens . sndLens :: Lens S Int
**head**of the list. - edits thetailLens = sndLens . sndLens :: Lens S [Int]
**tail**of the list.

Remarks:

- If the top level constructor of the list is , the head and the tail of the list can still be edited; the change will only be visible through[]when the constructor is changed back tolistLens. This may seem to be odd, but for many applications this is the right behaviour. See the interpretation subsection below.(:)
- For editing the tail of the tail of the list, we need an such thats' :: Sviewed throughsis the same astailLensviewed throughs'. Explained on a figure:listLens

*references*. One possible definition of references is described in LGtk/Semantics#References. How

### 3.2 Example: ADT with repeated record fields

Consider the following ADT:

data X a = X1 { y :: Int, z :: a } | X2 { y :: Int, v :: Char }

For the ADT lens, first define an auxiliary enum type for the constructor tags:

data XTag = X1Tag | X2Tag

xLens :: Lens (XTag, (Int, (a, Char))) (X a) xLens = lens get set where get (X1Tag, (y, (z, _))) = X1 y z get (X2Tag, (y, (_, v))) = X2 y v set (X1 y z) (_, (_, (_, v))) = (X1Tag, (y, (z, v))) set (X2 y v) (_, (_, (z, _))) = (X2Tag, (y, (z, v)))

Remarks:

- Instead of , we could use(XTag, (Int, (a, Char)))or(XTag, (Int, a, Char))too. This is an implementation detail.(XTag, Int, a, Char)
- remembers the value ofxLensif we change between the constructor tags. This is the intended behaviour.y
- remembers the values ofxLensandvfields if we change between the constructor tags. This is the intended behaviour.z

#### 3.2.1 Interpreation

The intended behaviour can be justified if we interpret lenses as abstract editors. If we would like to define an editor of an- An elementary editor for (maybe a combo box or a checkbox) which would be connected to the editor state byXTag.fstLens
- An elementary editor for an (maybe a text box or a slider) which would be connected to the editor state byInt.fstLens . sndLens
- An editor for an typed value which would be connected to the editor state bya.fstLens . sndLens . sndLens
- An editor for a (maybe a combo box or a text box or a virtual keyboard) which would be connected to the editor state byChar.fstLens . sndLens . sndLens . sndLens

Now, the intended behaviour is the following:

- If the user fills in an value forInt, this value should remain the same after changing theyvalue.XTag
- The value editor should be active only if theavalue isXTag.X1Tag
- The editor should be active only if theCharvalue isXTag.X2Tag
- If the user fills in an value forawhen thezvalue isXTag, and the user changesX1TagtoX1Tagand then back toX2Tag, theX1Tagvalue should be the same as before (consider a complex value which is hard to re-create). Similar holds for theavalue.Char

### 3.3 Generic ADT lens

In the generic case, consider the following ADT:

data X a1 ... aN = X1 { x11 :: t11, x12 :: t12, ... } | X2 { x21 :: t21, x22 :: t22, ... } ... | XM { xM1 :: tM1, xM2 :: tM2, ... }

Define an auxiliary enum type for the constructor tags:

data XTag = X1Tag | X2Tag | ... | XMTag

xLens :: Lens (XTag, (s1, (s2, ...))) (X a1 ... aN) xLens = lens get set where get (X1Tag, (s1, (s2, ...))) = X1 s{c11} s{c12} ... ... get (XMTag, (s1, (s2, ...))) = X2 s{cM1} s{cM2} ... -- replace second occurrence of the same variable name by _ in pattern set (X1 s{c11} s{c12} ...) (_, (s1, (s2, ...))) = (X1Tag, (s1, (s2, ...))) ... set (XM s{cM1} s{cM2} ...) (_, (s1, (s2, ...))) = (XMTag, (s1, (s2, ...)))

## 4 Summary

With the proposed ADT lenses, one can do arbitrary edit actions on ADT values in a user-friendly way (both distinct and common ADT parts are remembered when switching between constructors).

## 5 Links and references

I have not seen this technique described before. Please help to extend the list of papers / blog entries, where this or similar technique is used.