clojure.test.check.generators

any

A recursive generator that will generate many different, often nested, values

any-equatable

added in 0.10.0

Like any, but only generates objects that can be equal to other objects (e.g., do
not contain a NaN)

any-printable

Like any, but avoids characters that the shell will interpret as actions,
like 7 and 14 (bell and alternate character set command)

any-printable-equatable

added in 0.10.0

Like any, but avoids characters that the shell will interpret as actions,
like 7 and 14 (bell and alternate character set command), and only generates
objects that can be equal to other objects (e.g., do not contain a NaN)

big-ratio

added in 0.10.0

Generates a ratio (or integer) using gen/size-bounded-bigint. Shrinks
toward simpler ratios, which may be larger or smaller.

bind

(bind generator f)
Creates a new generator that passes the result of `gen` into function
`f`. `f` should return a new generator. This allows you to create new
generators that depend on the value of other generators. For example,
to create a generator of permutations which first generates a
`num-elements` and then generates a shuffling of `(range num-elements)`:

    (gen/bind gen/nat
              ;; this function takes a value generated by
              ;; the generator above and returns a new generator
              ;; which shuffles the collection returned by `range`
              (fn [num-elements]
                (gen/shuffle (range num-elements))))

Also see gen/let for a macro with similar functionality.

boolean

Generates one of `true` or `false`. Shrinks to `false`.

byte

Generates `java.lang.Byte`s, using the full byte-range.

bytes

Generates byte-arrays.

char

Generates character from 0-255.

char-alpha

Generates alpha characters.

char-alpha-numeric

deprecated in 0.6.0

Deprecated - use char-alphanumeric instead.

Generates alphanumeric characters.

char-alphanumeric

Generates alphanumeric characters.

char-ascii

Generates only ascii characters.

choose

(choose lower upper)
Creates a generator that generates integers uniformly in the range
`lower` to `upper`, inclusive.

    (gen/sample (gen/choose 200 800))
    => (331 241 593 339 643 718 688 473 247 694)

container-type

(container-type inner-type)

double

added in 0.9.0

Generates 64-bit floating point numbers from the entire range,
including +/- infinity and NaN. Use double* for more control.

double*

added in 0.9.0

(double* {:keys [infinite? NaN? min max], :or {infinite? true, NaN? true}})
Generates a 64-bit floating point number. Options:

  :infinite? - whether +/- infinity can be generated (default true)
  :NaN?      - whether NaN can be generated (default true)
  :min       - minimum value (inclusive, default none)
  :max       - maximum value (inclusive, default none)

Note that the min/max options must be finite numbers. Supplying a
min precludes -Infinity, and supplying a max precludes +Infinity.

elements

(elements coll)
Creates a generator that randomly chooses an element from `coll`.

(gen/sample (gen/elements [:foo :bar :baz]))
=> (:foo :baz :baz :bar :foo :foo :bar :bar :foo :bar)

fmap

(fmap f gen)
Returns a generator like `gen` but with values transformed by `f`.
E.g.:

    (gen/sample (gen/fmap str gen/nat))
    => ("0" "1" "0" "1" "4" "3" "6" "6" "4" "2")

Also see gen/let for a macro with similar functionality.

frequency

(frequency pairs)
Creates a generator that chooses a generator from `pairs` based on the
provided likelihoods. The likelihood of a given generator being chosen is
its likelihood divided by the sum of all likelihoods. Shrinks toward
choosing an earlier generator, as well as shrinking the value generated
by the chosen generator.

Examples:

    (gen/sample (gen/frequency [[5 gen/small-integer] [3 (gen/vector gen/small-integer)] [2 gen/boolean]]))
    => (true [] -1 [0] [1 -4 -4 1] true 4 [] 6 true)

generate

added in 0.8.0

(generate generator)(generate generator size)(generate generator size seed)
Returns a single sample value from the generator.

Note that this function is a dev helper and is not meant to be used
to build other generators.

Optional args:

- size: the abstract size parameter, defaults to 30
- seed: the seed for the random number generator, an integer

generator?

(generator? x)
Test if `x` is a generator. Generators should be treated as opaque values.

hash-map

(hash-map & kvs)
Like clojure.core/hash-map, except the values are generators.
Returns a generator that makes maps with the supplied keys and
values generated using the supplied generators.

    (gen/sample (gen/hash-map :a gen/boolean :b gen/nat))
    => ({:a false, :b 0}
        {:a true,  :b 1}
        {:a false, :b 2}
        {:a true,  :b 2}
        {:a false, :b 4}
        {:a false, :b 2}
        {:a true,  :b 3}
        {:a true,  :b 4}
        {:a false, :b 1}
        {:a false, :b 0})

int

deprecated in 0.10.0

Deprecated - use gen/small-integer instead.

Generates a positive or negative integer bounded by the generator's
`size` parameter.

keyword

Generates keywords without namespaces.

keyword-ns

added in 0.5.9

Generates keywords with namespaces.

large-integer

added in 0.9.0

Generates a platform-native integer from the full available range
(in clj, 64-bit Longs, and in cljs, numbers between -(2^53 - 1) and
(2^53 - 1)).

Use large-integer* for more control.

large-integer*

added in 0.9.0

(large-integer* {:keys [min max]})
Like large-integer, but accepts options:

  :min  the minimum integer (inclusive)
  :max  the maximum integer (inclusive)

Both :min and :max are optional.

    (gen/sample (gen/large-integer* {:min 9000 :max 10000}))
    => (9000 9001 9001 9002 9000 9003 9006 9030 9005 9044)

let

macro

added in 0.9.0

(let bindings & body)
Macro for building generators using values from other generators.
Uses a binding vector with the same syntax as clojure.core/let,
where the right-hand side of the binding pairs are generators, and
the left-hand side are names (or destructuring forms) for generated
values.

Subsequent generator expressions can refer to the previously bound
values, in the same way as clojure.core/let.

The body of the let can be either a value or a generator, and does
the expected thing in either case. In this way let provides the
functionality of both `bind` and `fmap`.

Examples:

  (gen/let [strs (gen/not-empty (gen/list gen/string))
            s (gen/elements strs)]
    {:some-strings strs
     :one-of-those-strings s})

  ;; generates collections of "users" that have integer IDs
  ;; from 0...N-1, but are in a random order
  (gen/let [users (gen/list (gen/hash-map :name gen/string-ascii
                                          :age gen/nat))]
    (->> users
         (map #(assoc %2 :id %1) (range))
         (gen/shuffle)))

list

(list generator)
Like `vector`, but generates lists.

list-distinct

added in 0.9.0

(list-distinct gen)(list-distinct gen opts)
Generates a list of elements from the given generator, with the
guarantee that the elements will be distinct.

If the generator cannot or is unlikely to produce enough distinct
elements, this generator will fail in the same way as `such-that`.

Available options:

  :num-elements  the fixed size of generated list
  :min-elements  the min size of generated list
  :max-elements  the max size of generated list
  :max-tries     the number of times the generator will be tried before
                 failing when it does not produce distinct elements
                 (default 10)
  :ex-fn         a function of one arg that will be called if test.check cannot
                 generate enough distinct values; it will be passed a map with
                 `:gen`, `:num-elements`, and `:max-tries` and should return an
                 exception

list-distinct-by

added in 0.9.0

(list-distinct-by key-fn gen)(list-distinct-by key-fn gen opts)
Generates a list of elements from the given generator, with the
guarantee that (map key-fn the-list) will be distinct.

If the generator cannot or is unlikely to produce enough distinct
elements, this generator will fail in the same way as `such-that`.

Available options:

  :num-elements  the fixed size of generated list
  :min-elements  the min size of generated list
  :max-elements  the max size of generated list
  :max-tries     the number of times the generator will be tried before
                 failing when it does not produce distinct elements
                 (default 10)
  :ex-fn         a function of one arg that will be called if test.check cannot
                 generate enough distinct values; it will be passed a map with
                 `:gen`, `:num-elements`, and `:max-tries` and should return an
                 exception

map

(map key-gen val-gen)(map key-gen val-gen opts)
Creates a generator that generates maps, with keys chosen from
`key-gen` and values chosen from `val-gen`.

If the key generator cannot or is unlikely to produce enough distinct
elements, this generator will fail in the same way as `such-that`.

Available options:

  :num-elements  the fixed size of generated maps
  :min-elements  the min size of generated maps
  :max-elements  the max size of generated maps
  :max-tries     the number of times the generator will be tried before
                 failing when it does not produce distinct elements
                 (default 10)
  :ex-fn         a function of one arg that will be called if test.check cannot
                 generate enough distinct keys; it will be passed a map with
                 `:gen` (the key-gen), `:num-elements`, and `:max-tries` and
                 should return an exception

nat

Generates non-negative integers bounded by the generator's `size`
parameter. Shrinks to zero.

neg-int

deprecated in 0.10.0

Deprecated - use (gen/fmap - gen/nat) instead (see also gen/large-integer).

(this generator, despite its name, can generate 0)

Generates nonpositive integers bounded by the generator's `size` parameter.

no-shrink

(no-shrink gen)
Creates a new generator that is just like `gen`, except does not shrink
at all. This can be useful when shrinking is taking a long time or is not
applicable to the domain.

not-empty

(not-empty gen)
Modifies a generator so that it doesn't generate empty collections.

Examples:

    ;; generate a vector of booleans, but never the empty vector
    (gen/sample (gen/not-empty (gen/vector gen/boolean)))
    => ([false]
        [false false]
        [false false]
        [false false false]
        [false false false false]
        [false true true]
        [true false false false]
        [true]
        [true true true false false true false]
        [false true true true false true true true false])

one-of

(one-of generators)
Creates a generator that randomly chooses a value from the list of
provided generators. Shrinks toward choosing an earlier generator,
as well as shrinking the value generated by the chosen generator.

    (gen/sample (gen/one-of [gen/small-integer gen/boolean (gen/vector gen/small-integer)]))
    => (true [] -1 [0] [1 -4 -4 1] true 4 [] 6 true)

pos-int

deprecated in 0.10.0

Deprecated - use gen/nat instead (see also gen/large-integer).

(this generator, despite its name, can generate 0)

Generates nonnegative integers bounded by the generator's `size` parameter.

ratio

Generates a small ratio (or integer) using gen/small-integer. Shrinks
toward simpler ratios, which may be larger or smaller.

recursive-gen

added in 0.5.9

(recursive-gen container-gen-fn scalar-gen)
This is a helper for writing recursive (tree-shaped) generators. The first
argument should be a function that takes a generator as an argument, and
produces another generator that 'contains' that generator. The vector function
in this namespace is a simple example. The second argument is a scalar
generator, like boolean. For example, to produce a tree of booleans:

  (gen/recursive-gen gen/vector gen/boolean)

Vectors or maps either recurring or containing booleans or integers:

  (gen/recursive-gen (fn [inner] (gen/one-of [(gen/vector inner)
                                              (gen/map inner inner)]))
                     (gen/one-of [gen/boolean gen/small-integer]))

Note that raw scalar values will be generated as well. To prevent this, you
can wrap the returned generator with the function passed as the first arg,
e.g.:

  (gen/vector (gen/recursive-gen gen/vector gen/boolean))

resize

(resize n generator)
Creates a new generator with `size` always bound to `n`.

(gen/sample (gen/set (gen/resize 200 gen/double)))
=> (#{}
    #{-4.994772362980037E147}
    #{-4.234418056487335E-146}
    #{}
    #{}
    #{}
    #{NaN}
    #{8.142414100982609E-63}
    #{-3.58429955903876E-159 2.8563794617604296E-154
      4.1021360195776005E-100 1.9084564045332549E-38}
    #{-2.1582818131881376E83 -5.8460065493236117E48 9.729260993803226E166})

return

(return value)
Creates a generator that always returns `value`,
and never shrinks. You can think of this as
the `constantly` of generators. E.g.:

    (gen/sample (gen/return 42))
    => (42 42 42 42 42 42 42 42 42 42)

s-neg-int

deprecated in 0.10.0

Deprecated - use (gen/fmap (comp dec -) gen/nat) instead (see also gen/large-integer).

Generates negative integers bounded by the generator's `size` + 1

s-pos-int

deprecated in 0.10.0

Deprecated - use (gen/fmap inc gen/nat) instead (see also gen/large-integer).

Generates positive integers bounded by the generator's `size` + 1

sample

(sample generator)(sample generator num-samples)
Return a sequence of `num-samples` (default 10)
realized values from `generator`.

The sequence starts with small values from the generator, which
probably do not reflect the variety of values that will be generated
during a longer test run.

Note that this function is a dev helper and is not meant to be used
to build other generators.

sample-seq

(sample-seq generator)(sample-seq generator max-size)
Returns an infinite sequence of realized values from `generator`.

Note that this function is a dev helper and is not meant to be used
to build other generators.

scale

added in 0.8.0

(scale f generator)
Creates a new generator that modifies the size parameter by the
given function. Intended to support generators with sizes that need
to grow at different rates compared to the normal linear scaling.

    (gen/sample (gen/tuple (gen/scale #(/ % 10) gen/nat)
                           gen/nat
                           (gen/scale #(* % 10) gen/nat)))
    => ([0 0 0]  [0 1 2]  [0 2 13] [0 1 6]  [0 1 23]
        [0 2 42] [0 1 26] [0 1 12] [0 1 12] [0 0 3])

set

added in 0.9.0

(set gen)(set gen opts)
Generates a set of elements from the given generator.

If the generator cannot or is unlikely to produce enough distinct
elements, this generator will fail in the same way as `such-that`.

Available options:

  :num-elements  the fixed size of generated set
  :min-elements  the min size of generated set
  :max-elements  the max size of generated set
  :max-tries     the number of times the generator will be tried before
                 failing when it does not produce distinct elements
                 (default 10)
  :ex-fn         a function of one arg that will be called if test.check cannot
                 generate enough distinct values; it will be passed a map with
                 `:gen`, `:num-elements`, and `:max-tries` and should return an
                 exception

shrink-2

(shrink-2 gen)
Creates a new generator like `gen`, but will consider nodes for shrinking
even if their parent passes the test (up to one additional level).

shuffle

added in 0.6.0

(shuffle coll)
Creates a generator that generates random permutations of
`coll`. Shrinks toward the original collection: `coll`. `coll` will
be coerced to a vector.

simple-type

Generates a variety of scalar types.

simple-type-equatable

added in 0.10.0

Like gen/simple-type, but only generates objects that can be
equal to other objects (e.g., not a NaN).

simple-type-printable

Generates a variety of scalar types, with printable strings.

simple-type-printable-equatable

added in 0.10.0

Like gen/simple-type-printable, but only generates objects that
can be equal to other objects (e.g., not a NaN).

size-bounded-bigint

added in 0.10.0

Generates an integer (long or bigint) bounded exclusively by ±2^(6*size).

sized

(sized sized-gen)
Creates a generator that depends on the size parameter.
`sized-gen` is a function that takes an integer and returns
a generator.

Examples:

    ;; generates vectors of booleans where the length always exactly
    ;; matches the `size` parameter
    (gen/sample (gen/sized (fn [size] (gen/vector gen/boolean size))))
    => ([]
        [false]
        [true true]
        [false true false]
        [false true true true]
        [false false true true false]
        [false true false true true false]
        [true false true true true false false]
        [true true false false false true false false]
        [false false false true true false true false true])

small-integer

added in 0.10.0

Generates a positive or negative integer bounded by the generator's
`size` parameter. Shrinks to zero.

sorted-set

added in 0.9.0

(sorted-set gen)(sorted-set gen opts)
Generates a sorted set of elements from the given generator.

If the generator cannot or is unlikely to produce enough distinct
elements, this generator will fail in the same way as `such-that`.

Available options:

  :num-elements  the fixed size of generated set
  :min-elements  the min size of generated set
  :max-elements  the max size of generated set
  :max-tries     the number of times the generator will be tried before
                 failing when it does not produce distinct elements
                 (default 10)
  :ex-fn         a function of one arg that will be called if test.check cannot
                 generate enough distinct values; it will be passed a map with
                 `:gen`, `:num-elements`, and `:max-tries` and should return an
                 exception

string

Generates strings. May generate unprintable characters.

string-alpha-numeric

deprecated in 0.6.0

Deprecated - use string-alphanumeric instead.

Generates alphanumeric strings.

string-alphanumeric

Generates alphanumeric strings.

string-ascii

Generates ascii strings.

such-that

(such-that pred gen)(such-that pred gen max-tries-or-opts)
Creates a generator that generates values from `gen` that satisfy predicate
`pred`. Care is needed to ensure there is a high chance `gen` will satisfy
`pred`. By default, `such-that` will try 10 times to generate a value that
satisfies the predicate. If no value passes this predicate after this number
of iterations, a runtime exception will be thrown. Note also that each
time such-that retries, it will increase the size parameter.

Examples:

    ;; generate non-empty vectors of integers
    ;; (note, gen/not-empty does exactly this)
    (gen/such-that not-empty (gen/vector gen/small-integer))

You can customize `such-that` by passing an optional third argument, which can
either be an integer representing the maximum number of times test.check
will try to generate a value matching the predicate, or a map:

    :max-tries  positive integer, the maximum number of tries (default 10)
    :ex-fn      a function of one arg that will be called if test.check cannot
                generate a matching value; it will be passed a map with `:gen`,
                `:pred`, and `:max-tries` and should return an exception

symbol

Generates symbols without namespaces.

symbol-ns

added in 0.5.9

Generates symbols with namespaces.

tuple

(tuple & generators)
Creates a generator that returns a vector, whose elements are chosen
from the generators in the same position. The individual elements shrink
according to their generator, but the vector will never shrink in count.

Examples:

    (def t (tuple gen/small-integer gen/boolean))
    (sample t)
    ;; => ([1 true] [2 true] [2 false] [1 false] [0 true] [-2 false] [-6 false]
    ;; =>  [3 true] [-4 false] [9 true]))

uuid

added in 0.9.0

Generates a random type-4 UUID. Does not shrink.

vector

(vector generator)(vector generator num-elements)(vector generator min-elements max-elements)
Creates a generator of vectors whose elements are chosen from
`generator`. The count of the vector will be bounded by the `size`
generator parameter.

vector-distinct

added in 0.9.0

(vector-distinct gen)(vector-distinct gen opts)
Generates a vector of elements from the given generator, with the
guarantee that the elements will be distinct.

If the generator cannot or is unlikely to produce enough distinct
elements, this generator will fail in the same way as `such-that`.

Available options:

  :num-elements  the fixed size of generated vectors
  :min-elements  the min size of generated vectors
  :max-elements  the max size of generated vectors
  :max-tries     the number of times the generator will be tried before
                 failing when it does not produce distinct elements
                 (default 10)
  :ex-fn         a function of one arg that will be called if test.check cannot
                 generate enough distinct values; it will be passed a map with
                 `:gen`, `:num-elements`, and `:max-tries` and should return an
                 exception

vector-distinct-by

added in 0.9.0

(vector-distinct-by key-fn gen)(vector-distinct-by key-fn gen opts)
Generates a vector of elements from the given generator, with the
guarantee that (map key-fn the-vector) will be distinct.

If the generator cannot or is unlikely to produce enough distinct
elements, this generator will fail in the same way as `such-that`.

Available options:

  :num-elements  the fixed size of generated vectors
  :min-elements  the min size of generated vectors
  :max-elements  the max size of generated vectors
  :max-tries     the number of times the generator will be tried before
                 failing when it does not produce distinct elements
                 (default 10)
  :ex-fn         a function of one arg that will be called if test.check cannot
                 generate enough distinct values; it will be passed a map with
                 `:gen`, `:num-elements`, and `:max-tries` and should return an
                 exception