Sanctuary

Sanctuary is a small functional programming library inspired by Haskell and PureScript. It depends on and works nicely with Ramda. Sanctuary makes it possible to write safe code without null checks.

In JavaScript it's trivial to introduce a possible run-time type error:

words[0].toUpperCase()

If words is [] we'll get a familiar error at run-time:

TypeError: Cannot read property 'toUpperCase' of undefined

Sanctuary gives us a fighting chance of avoiding such errors. We might write:

R.map(R.toUpper, S.head(words))

Types

Sanctuary uses Haskell-like type signatures to describe the types of values, including functions. 'foo', for example, has type String; [1, 2, 3] has type [Number]. The arrow (->) is used to express a function's type. Math.abs, for example, has type Number -> Number. That is, it takes an argument of type Number and returns a value of type Number.

R.map has type (a -> b) -> [a] -> [b]. That is, it takes an argument of type a -> b and returns a value of type [a] -> [b]. a and b are type variables: applying R.map to a value of type String -> Number will give a value of type [String] -> [Number].

Sanctuary embraces types. JavaScript doesn't support algebraic data types, but these can be simulated by providing a group of constructor functions whose prototypes provide the same set of methods. A value of the Maybe type, for example, is created via the Nothing constructor or the Just constructor.

It's necessary to extend Haskell's notation to describe implicit arguments to the methods provided by Sanctuary's types. In x.map(y), for example, the map method takes an implicit argument x in addition to the explicit argument y. The type of the value upon which a method is invoked appears at the beginning of the signature, separated from the arguments and return value by a squiggly arrow (~>). The type of the map method of the Maybe type is written Maybe a ~> (a -> b) -> Maybe b. One could read this as:

When the map method is invoked on a value of type Maybe a (for any type a) with an argument of type a -> b (for any type b), it returns a value of type Maybe b.

API

Combinator

`K :: a -> b -> a`

The K combinator. Takes two values and returns the first. Equivalent to Haskell's const function.

> S.K('foo', 'bar')
"foo"
> R.map(S.K(42), R.range(0, 5))
[42, 42, 42, 42, 42]

Maybe type

`Maybe :: Type`

The Maybe type represents optional values: a value of type Maybe a is either a Just whose value is of type a or a Nothing (with no value).

The Maybe type satisfies the Monoid and Monad specifications.

`Maybe.empty :: -> Maybe a`

Returns a Nothing.

> S.Maybe.empty()
Nothing()

`Maybe.of :: a -> Maybe a`

Takes a value of any type and returns a Just with the given value.

> S.Maybe.of(42)
Just(42)

`Maybe#ap :: Maybe (a -> b) ~> Maybe a -> Maybe b`

Takes a value of type Maybe a and returns a Nothing unless this is a Just and the argument is a Just, in which case it returns a Just whose value is the result of of applying this Just's value to the given Just's value.

> S.Nothing().ap(S.Just(42))
Nothing()

> S.Just(R.inc).ap(S.Nothing())
Nothing()

> S.Just(R.inc).ap(S.Just(42))
Just(43)

`Maybe#chain :: Maybe a ~> (a -> Maybe b) -> Maybe b`

Takes a function and returns this if this is a Nothing; otherwise it returns the result of applying the function to this Just's value.

> S.Nothing().chain(S.parseFloat)
Nothing()

> S.Just('xxx').chain(S.parseFloat)
Nothing()

> S.Just('12.34').chain(S.parseFloat)
Just(12.34)

`Maybe#concat :: Maybe a ~> Maybe a -> Maybe a`

Returns the result of concatenating two Maybe values of the same type. a must have a Semigroup (indicated by the presence of a concat method).

If this is a Nothing and the argument is a Nothing, this method returns a Nothing.

If this is a Just and the argument is a Just, this method returns a Just whose value is the result of concatenating this Just's value and the given Just's value.

Otherwise, this method returns the Just.

> S.Nothing().concat(S.Nothing())
Nothing()

> S.Just([1, 2, 3]).concat(S.Just([4, 5, 6]))
Just([1, 2, 3, 4, 5, 6])

> S.Nothing().concat(S.Just([1, 2, 3]))
Just([1, 2, 3])

> S.Just([1, 2, 3]).concat(S.Nothing())
Just([1, 2, 3])

`Maybe#empty :: Maybe a ~> Maybe a`

Returns a Nothing.

> S.Just(42).empty()
Nothing()

`Maybe#equals :: Maybe a ~> b -> Boolean`

Takes a value of any type and returns true if:

it is a Nothing and this is a Nothing; or
it is a Just and this is a Just, and their values are equal according to R.equals.

> S.Nothing().equals(S.Nothing())
true

> S.Nothing().equals(null)
false

> S.Just([1, 2, 3]).equals(S.Just([1, 2, 3]))
true

> S.Just([1, 2, 3]).equals(S.Just([3, 2, 1]))
false

> S.Just([1, 2, 3]).equals(S.Nothing())
false

`Maybe#filter :: Maybe a ~> (a -> Boolean) -> Maybe a`

Takes a predicate and returns this if this is a Just whose value satisfies the predicate; Nothing otherwise.

> S.Just(42).filter(function(n) { return n % 2 === 0; })
Just(42)

> S.Just(43).filter(function(n) { return n % 2 === 0; })
Nothing()

`Maybe#map :: Maybe a ~> (a -> b) -> Maybe b`

Takes a function and returns this if this is a Nothing; otherwise it returns a Just whose value is the result of applying the function to this Just's value.

> S.Nothing().map(R.inc)
Nothing()

> S.Just(42).map(R.inc)
Just(43)

`Maybe#of :: Maybe a ~> b -> Maybe b`

Takes a value of any type and returns a Just with the given value.

> S.Nothing().of(42)
Just(42)

`Maybe#toBoolean :: Maybe a ~> Boolean`

Returns false if this is a Nothing; true if this is a Just.

> S.Nothing().toBoolean()
false

> S.Just(42).toBoolean()
true

`Maybe#toString :: Maybe a ~> String`

Returns the string representation of the Maybe.

> S.Nothing().toString()
"Nothing()"

> S.Just([1, 2, 3]).toString()
"Just([1, 2, 3])"

`Maybe#type :: Type`

A reference to the Maybe type. Useful for determining whether two values such as S.Nothing() and S.Just(42) are of the same type.

`Nothing :: -> Maybe a`

Returns a Nothing. Though this is a constructor function the new keyword needn't be used.

> S.Nothing()
Nothing()

`Just :: a -> Maybe a`

Takes a value of any type and returns a Just with the given value. Though this is a constructor function the new keyword needn't be used.

> S.Just(42)
Just(42)

`fromMaybe :: a -> Maybe a -> a`

Takes a default value and a Maybe, and returns the Maybe's value if the Maybe is a Just; the default value otherwise.

> S.fromMaybe(0, S.Just(42))
42

> S.fromMaybe(0, S.Nothing())
0

`toMaybe :: a? -> Maybe a`

Takes a value and returns Nothing if the value is null or undefined; Just the value otherwise.

> S.toMaybe(null)
Nothing()

> S.toMaybe(42)
Just(42)

`encase :: (* -> a) -> (* -> Maybe a)`

Takes a function f which may throw and returns a curried function g which will not throw. The result of applying g is determined by applying f to the same arguments: if this succeeds, g returns Just the result; otherwise g returns Nothing.

> S.encase(eval)('1 + 1')
Just(2)

> S.encase(eval)('1 +')
Nothing()

Either type

`Either :: Type`

The Either type represents values with two possibilities: a value of type Either a b is either a Left whose value is of type a or a Right whose value is of type b.

The Either type satisfies the Semigroup and Monad specifications.

`Either.of :: b -> Either a b`

Takes a value of any type and returns a Right with the given value.

> S.Either.of(42)
Right(42)

`Either#ap :: Either a (b -> c) ~> Either a b -> Either a c`

Takes a value of type Either a b and returns a Left unless this is a Right and the argument is a Right, in which case it returns a Right whose value is the result of applying this Right's value to the given Right's value.

> S.Left('Cannot divide by zero').ap(S.Right(42))
Left("Cannot divide by zero")

> S.Right(R.inc).ap(S.Left('Cannot divide by zero'))
Left("Cannot divide by zero")

> S.Right(R.inc).ap(S.Right(42))
Right(43)

`Either#chain :: Either a b ~> (b -> Either a c) -> Either a c`

Takes a function and returns this if this is a Left; otherwise it returns the result of applying the function to this Right's value.

> void (sqrt = function(n) { return n < 0 ? S.Left('Cannot represent square root of negative number') : S.Right(Math.sqrt(n)); })
undefined

> S.Left('Cannot divide by zero').chain(sqrt)
Left("Cannot divide by zero")

> S.Right(-1).chain(sqrt)
Left("Cannot represent square root of negative number")

> S.Right(25).chain(sqrt)
Right(5)

`Either#concat :: Either a b ~> Either a b -> Either a b`

Returns the result of concatenating two Either values of the same type. a must have a Semigroup (indicated by the presence of a concat method), as must b.

If this is a Left and the argument is a Left, this method returns a Left whose value is the result of concatenating this Left's value and the given Left's value.

If this is a Right and the argument is a Right, this method returns a Right whose value is the result of concatenating this Right's value and the given Right's value.

Otherwise, this method returns the Right.

> S.Left('abc').concat(S.Left('def'))
Left("abcdef")

> S.Right([1, 2, 3]).concat(S.Right([4, 5, 6]))
Right([1, 2, 3, 4, 5, 6])

> S.Left('abc').concat(S.Right([1, 2, 3]))
Right([1, 2, 3])

> S.Right([1, 2, 3]).concat(S.Left('abc'))
Right([1, 2, 3])

`Either#equals :: Either a b ~> c -> Boolean`

Takes a value of any type and returns true if:

it is a Left and this is a Left, and their values are equal according to R.equals; or
it is a Right and this is a Right, and their values are equal according to R.equals.

> S.Right([1, 2, 3]).equals(S.Right([1, 2, 3]))
true

> S.Right([1, 2, 3]).equals(S.Left([1, 2, 3]))
false

> S.Right(42).equals(42)
false

`Either#map :: Either a b ~> (b -> c) -> Either a c`

Takes a function and returns this if this is a Left; otherwise it returns a Right whose value is the result of applying the function to this Right's value.

> S.Left('Cannot divide by zero').map(R.inc)
Left("Cannot divide by zero")

> S.Right(42).map(R.inc)
Right(43)

`Either#of :: Either a b ~> b -> Either a b`

Takes a value of any type and returns a Right with the given value.

> S.Left('Cannot divide by zero').of(42)
Right(42)

`Either#toBoolean :: Either a b ~> Boolean`

Returns false if this is a Left; true if this is a Right.

> S.Left(42).toBoolean()
false

> S.Right(42).toBoolean()
true

`Either#toString :: Either a b ~> String`

Returns the string representation of the Either.

> S.Left('Cannot divide by zero').toString()
"Left(\\"Cannot divide by zero\\")"

> S.Right([1, 2, 3]).toString()
"Right([1, 2, 3])"

`Either#type :: Type`

A reference to the Either type. Useful for determining whether two values such as S.Left('Cannot divide by zero') and S.Right(42) are of the same type.

`Left :: a -> Either a b`

Takes a value of any type and returns a Left with the given value. Though this is a constructor function the new keyword needn't be used.

> S.Left('Cannot divide by zero')
Left("Cannot divide by zero")

`Right :: b -> Either a b`

Takes a value of any type and returns a Right with the given value. Though this is a constructor function the new keyword needn't be used.

> S.Right(42)
Right(42)

`either :: (a -> c) -> (b -> c) -> Either a b -> c`

Takes two functions and an Either, and returns the result of applying the first function to the Left's value, if the Either is a Left, or the result of applying the second function to the Right's value, if the Either is a Right.

> S.either(R.toUpper, R.toString, S.Left('Cannot divide by zero'))
"CANNOT DIVIDE BY ZERO"

> S.either(R.toUpper, R.toString, S.Right(42))
"42"

Control

`and :: a -> a -> a`

Takes two values of the same type and returns the second value if the first is "true"; the first value otherwise. An array is considered "true" if its length is greater than zero. The Boolean value true is also considered "true". Other types must provide a toBoolean method.

> S.and(S.Just(1), S.Just(2))
Just(2)

> S.and(S.Nothing(), S.Just(3))
Nothing()

`or :: a -> a -> a`

Takes two values of the same type and returns the first value if it is "true"; the second value otherwise. An array is considered "true" if its length is greater than zero. The Boolean value true is also considered "true". Other types must provide a toBoolean method.

> S.or(S.Just(1), S.Just(2))
Just(1)

> S.or(S.Nothing(), S.Just(3))
Just(3)

`xor :: a -> a -> a`

Takes two values of the same type and returns the "true" value if one value is "true" and the other is "false"; otherwise it returns the type's "false" value. An array is considered "true" if its length is greater than zero. The Boolean value true is also considered "true". Other types must provide toBoolean and empty methods.

> S.xor(S.Nothing(), S.Just(1))
Just(1)

> S.xor(S.Just(2), S.Just(3))
Nothing()

List

`slice :: Number -> Number -> [a] -> Maybe [a]`

Returns Just a list containing the elements from the supplied list from a beginning index (inclusive) to an end index (exclusive). Returns Nothing unless the start interval is less than or equal to the end interval, and the list contains both (half-open) intervals. Accepts negative indices, which indicate an offset from the end of the list.

Dispatches to its third argument's slice method if present. As a result, one may replace [a] with String in the type signature.

> S.slice(1, 3, ['a', 'b', 'c', 'd', 'e'])
Just(["b", "c"])

> S.slice(-2, -0, ['a', 'b', 'c', 'd', 'e'])
Just(["d", "e"])

> S.slice(2, -0, ['a', 'b', 'c', 'd', 'e'])
Just(["c", "d", "e"])

> S.slice(1, 6, ['a', 'b', 'c', 'd', 'e'])
Nothing()

> S.slice(2, 6, 'banana')
Just("nana")

`at :: Number -> [a] -> Maybe a`

Takes an index and a list and returns Just the element of the list at the index if the index is within the list's bounds; Nothing otherwise. A negative index represents an offset from the length of the list.

> S.at(2, ['a', 'b', 'c', 'd', 'e'])
Just("c")

> S.at(5, ['a', 'b', 'c', 'd', 'e'])
Nothing()

> S.at(-2, ['a', 'b', 'c', 'd', 'e'])
Just("d")

`head :: [a] -> Maybe a`

Takes a list and returns Just the first element of the list if the list contains at least one element; Nothing if the list is empty.

> S.head([1, 2, 3])
Just(1)

> S.head([])
Nothing()

`last :: [a] -> Maybe a`

Takes a list and returns Just the last element of the list if the list contains at least one element; Nothing if the list is empty.

> S.last([1, 2, 3])
Just(3)

> S.last([])
Nothing()

`tail :: [a] -> Maybe [a]`

Takes a list and returns Just a list containing all but the first of the list's elements if the list contains at least one element; Nothing if the list is empty.

> S.tail([1, 2, 3])
Just([2, 3])

> S.tail([])
Nothing()

`init :: [a] -> Maybe [a]`

Takes a list and returns Just a list containing all but the last of the list's elements if the list contains at least one element; Nothing if the list is empty.

> S.init([1, 2, 3])
Just([1, 2])

> S.init([])
Nothing()

`take :: Number -> [a] -> Maybe [a]`

Returns Just the first N elements of the given collection if N is greater than or equal to zero and less than or equal to the length of the collection; Nothing otherwise. Supports Array, String, and any other collection type which provides a slice method.

> S.take(2, ['a', 'b', 'c', 'd', 'e'])
Just(["a", "b"])

> S.take(4, 'abcdefg')
Just("abcd")

> S.take(4, ['a', 'b', 'c'])
Nothing()

`drop :: Number -> [a] -> Maybe [a]`

Returns Just all but the first N elements of the given collection if N is greater than or equal to zero and less than or equal to the length of the collection; Nothing otherwise. Supports Array, String, and any other collection type which provides a slice method.

> S.drop(2, ['a', 'b', 'c', 'd', 'e'])
Just(["c", "d", "e"])

> S.drop(4, 'abcdefg')
Just("efg")

> S.drop(4, 'abc')
Nothing()

`find :: (a -> Boolean) -> [a] -> Maybe a`

Takes a predicate and a list and returns Just the leftmost element of the list which satisfies the predicate; Nothing if none of the list's elements satisfies the predicate.

> S.find(function(n) { return n < 0; }, [1, -2, 3, -4, 5])
Just(-2)

> S.find(function(n) { return n < 0; }, [1, 2, 3, 4, 5])
Nothing()

`indexOf :: a -> [a] -> Maybe Number`

Takes a value of any type and a list, and returns Just the index of the first occurrence of the value in the list, if applicable; Nothing otherwise.

Dispatches to its second argument's indexOf method if present. As a result, String -> String -> Maybe Number is an alternative type signature.

> S.indexOf('a', ['b', 'a', 'n', 'a', 'n', 'a'])
Just(1)

> S.indexOf('x', ['b', 'a', 'n', 'a', 'n', 'a'])
Nothing()

> S.indexOf('an', 'banana')
Just(1)

> S.indexOf('ax', 'banana')
Nothing()

`lastIndexOf :: a -> [a] -> Maybe Number`

Takes a value of any type and a list, and returns Just the index of the last occurrence of the value in the list, if applicable; Nothing otherwise.

Dispatches to its second argument's lastIndexOf method if present. As a result, String -> String -> Maybe Number is an alternative type signature.

> S.lastIndexOf('a', ['b', 'a', 'n', 'a', 'n', 'a'])
Just(5)

> S.lastIndexOf('x', ['b', 'a', 'n', 'a', 'n', 'a'])
Nothing()

> S.lastIndexOf('an', 'banana')
Just(3)

> S.lastIndexOf('ax', 'banana')
Nothing()

`pluck :: String -> [{String: }] -> [Maybe ]`

Takes a list of objects and plucks the value of the specified key for each object in the list. Returns the value wrapped in a Just if an object has the key and a Nothing if it does not.

> S.pluck('a', [{a: 1, b: 2}, {a: 4, b: 5}, {b: 3, c: 7}])
[Just(1), Just(4), Nothing()]

> S.pluck('x', [{x: 1}, {x: 2}, {x: undefined}])
[Just(1), Just(2), Just(undefined)]

Object

`get :: String -> Object -> Maybe *`

Takes a property name and an object and returns Just the value of the specified property of the object if the object has such an own property; Nothing otherwise.

> S.get('x', {x: 1, y: 2})
Just(1)

> S.get('toString', {x: 1, y: 2})
Nothing()

`gets :: [String] -> Object -> Maybe *`

Takes a list of property names and an object and returns Just the value at the path specified by the list of property names if such a path exists; Nothing otherwise.

> S.gets(['a', 'b', 'c'], {a: {b: {c: 42}}})
Just(42)

> S.gets(['a', 'b', 'c'], {})
Nothing()

Parse

`parseDate :: String -> Maybe Date`

Takes a string and returns Just the date represented by the string if it does in fact represent a date; Nothing otherwise.

> S.parseDate('2011-01-19T17:40:00Z')
Just(new Date("2011-01-19T17:40:00.000Z"))

> S.parseDate('today')
Nothing()

`parseFloat :: String -> Maybe Number`

Takes a string and returns Just the number represented by the string if it does in fact represent a number; Nothing otherwise.

> S.parseFloat('-123.45')
Just(-123.45)

> S.parseFloat('foo.bar')
Nothing()

`parseInt :: Number -> String -> Maybe Number`

Takes a radix (an integer between 2 and 36 inclusive) and a string, and returns Just the number represented by the string if it does in fact represent a number in the base specified by the radix; Nothing otherwise.

This function is stricter than parseInt: a string is considered to represent an integer only if all its non-prefix characters are members of the character set specified by the radix.

> S.parseInt(10, '-42')
Just(-42)

> S.parseInt(16, '0xFF')
Just(255)

> S.parseInt(16, '0xGG')
Nothing()

`parseJson :: String -> Maybe *`

Takes a string which may or may not be valid JSON, and returns Just the result of applying JSON.parse to the string if valid; Nothing otherwise.

> S.parseJson('["foo","bar","baz"]')
Just(["foo", "bar", "baz"])

> S.parseJson('[')
Nothing()

RegExp

`match :: RegExp -> String -> Maybe [Maybe String]`

Takes a pattern and a string, and returns Just a list of matches if the pattern matches the string; Nothing otherwise. Each match has type Maybe String, where a Nothing represents an unmatched optional capturing group.

> S.match(/(good)?bye/, 'goodbye')
Just([Just("goodbye"), Just("good")])

> S.match(/(good)?bye/, 'bye')
Just([Just("bye"), Nothing()])