# array

## Functional, extendible arrays.

Functional, extendible arrays. Arrays can have fixed size, or can grow automatically as needed. A default value is used for entries that have not been explicitly set.

Arrays uses *zero* based indexing. This is a deliberate design
choice and differs from other erlang datastructures, e.g. tuples.

Unless specified by the user when the array is created, the default
value is the atom `undefined`

. There is no difference between an
unset entry and an entry which has been explicitly set to the same
value as the default one (cf. reset/2). If you need to
differentiate between unset and set entries, you must make sure that
the default value cannot be confused with the values of set entries.

The array never shrinks automatically; if an index `I`

has been used
successfully to set an entry, all indices in the range [0,`I`

] will
stay accessible unless the array size is explicitly changed by
calling resize/2.

Examples:

%% Create a fixed-size array with entries 0-9 set to 'undefined' A0 = array:new(10). 10 = array:size(A0). %% Create an extendible array and set entry 17 to 'true', %% causing the array to grow automatically A1 = array:set(17, true, array:new()). 18 = array:size(A1). %% Read back a stored value true = array:get(17, A1). %% Accessing an unset entry returns the default value undefined = array:get(3, A1). %% Accessing an entry beyond the last set entry also returns the %% default value, if the array does not have fixed size undefined = array:get(18, A1). %% "sparse" functions ignore default-valued entries A2 = array:set(4, false, A1). [{4, false}, {17, true}] = array:sparse_to_orddict(A2). %% An extendible array can be made fixed-size later A3 = array:fix(A2). %% A fixed-size array does not grow automatically and does not %% allow accesses beyond the last set entry {'EXIT',{badarg,_}} = (catch array:set(18, true, A3)). {'EXIT',{badarg,_}} = (catch array:get(18, A3)).

#### DATA TYPES

`array()`

A functional, extendible array. The representation is not documented and is subject to change without notice. Note that arrays cannot be directly compared for equality.

#### Functions

### default(Array::array()) -> term()

### fix(Array::array()) -> array()

### foldl(Function, InitialAcc::term(), Array::array()) -> term()

`Function = (Index::integer(), Value::term(), Acc::term()) -> term()`

Fold the elements of the array using the given function and
initial accumulator value. The elements are visited in order from the
lowest index to the highest. If `Function`

is not a function, the
call fails with reason `badarg`

.

*See also:* foldr/3, map/2, sparse_foldl/3.

### foldr(Function, InitialAcc::term(), Array::array()) -> term()

`Function = (Index::integer(), Value::term(), Acc::term()) -> term()`

### from_list(List::list()) -> array()

Equivalent to from_list(List, undefined).

### from_list(List::list(), Default::term()) -> array()

### from_orddict(Orddict::list()) -> array()

Equivalent to from_orddict(Orddict, undefined).

### from_orddict(List::list(), Default::term()) -> array()

Convert an ordered list of pairs `{Index, Value}`

to a
corresponding extendible array. `Default`

is used as the value for
uninitialized entries of the array. If `List`

is not a proper,
ordered list of pairs whose first elements are nonnegative
integers, the call fails with reason `badarg`

.

*See also:* new/2, to_orddict/1.

### get(I::integer(), Array::array()) -> term()

Get the value of entry `I`

. If `I`

is not a nonnegative
integer, or if the array has fixed size and `I`

is larger than the
maximum index, the call fails with reason `badarg`

.

If the array does not have fixed size, this function will return the
default value for any index `I`

greater than `size(Array)-1`

.

*See also:* set/3.

### is_array(X::term()) -> bool()

Returns `true`

if `X`

appears to be an array, otherwise `false`

.
Note that the check is only shallow; there is no guarantee that `X`

is a well-formed array representation even if this function returns
`true`

.

### is_fix(Array::array()) -> bool()

Check if the array has fixed size.
Returns `true`

if the array is fixed, otherwise `false`

.

*See also:* fix/1.

### map(Function, Array::array()) -> array()

`Function = (Index::integer(), Value::term()) -> term()`

Map the given function onto each element of the array. The
elements are visited in order from the lowest index to the highest.
If `Function`

is not a function, the call fails with reason `badarg`

.

*See also:* foldl/3, foldr/3, sparse_map/2.

### new(Options::term()) -> array()

Create a new array according to the given options. By default, the array is extendible and has initial size zero. Array indices start at 0.

`Options`

is a single term or a list of terms, selected from the
following:

`N::integer()`

or `{size, N::integer()}`

Specifies the initial size of the array; this also implies
`{fixed, true}`

. If `N`

is not a nonnegative integer, the call
fails with reason `badarg`

.

`fixed`

or `{fixed, true}`

Creates a fixed-size array; see also fix/1.

`{fixed, false}`

Creates an extendible (non fixed-size) array.

`{default, Value}`

Sets the default value for the array to `Value`

.

Options are processed in the order they occur in the list, i.e., later options have higher precedence.

The default value is used as the value of uninitialized entries, and cannot be changed once the array has been created.

Examples:

array:new(100)

creates a fixed-size array of size 100.

array:new({default,0})

creates an empty, extendible array whose default value is 0.

array:new([{size,10},{fixed,false},{default,-1}])

creates an extendible array with initial size 10 whose default value is -1.

### new(Size::integer(), Options::term()) -> array()

Create a new array according to the given size and options. If
`Size`

is not a nonnegative integer, the call fails with reason
`badarg`

. By default, the array has fixed size. Note that any size
specifications in `Options`

will override the `Size`

parameter.

If `Options`

is a list, this is simply equivalent to ```
new([{size,
Size} | Options]
```

, otherwise it is equivalent to ```
new([{size, Size} |
[Options]]
```

. However, using this function directly is more efficient.

Example:

array:new(100, {default,0})

creates a fixed-size array of size 100, whose default value is 0.

*See also:* new/1.

### relax(Array::array()) -> array()

### reset(I::integer(), Array::array()) -> array()

Reset entry `I`

to the default value for the array. This is
equivalent to `set(I, default(Array), Array)`

, and hence may cause
the array to grow in size, but will not shrink it. Shrinking can be
done explicitly by calling resize/2.

If `I`

is not a nonnegative integer, or if the array has fixed size
and `I`

is larger than the maximum index, the call fails with reason
`badarg`

; cf. set/3

### resize(Array::array()) -> array()

Change the size of the array to that reported by sparse_size/1. If the given array has fixed size, the resulting array will also have fixed size.

*See also:* resize/2, sparse_size/1.

### resize(Size::integer(), Array::array()) -> array()

Change the size of the array. If `Size`

is not a nonnegative
integer, the call fails with reason `badarg`

. If the given array has
fixed size, the resulting array will also have fixed size.

### set(I::integer(), Value::term(), Array::array()) -> array()

Set entry `I`

of the array to `Value`

. If `I`

is not a
nonnegative integer, or if the array has fixed size and `I`

is larger
than the maximum index, the call fails with reason `badarg`

.

If the array does not have fixed size, and `I`

is greater than
`size(Array)-1`

, the array will grow to size `I+1`

.

### size(Array::array()) -> integer()

Get the number of entries in the array. Entries are numbered
from 0 to `size(Array)-1`

; hence, this is also the index of the first
entry that is guaranteed to not have been previously set.

*See also:* set/3, sparse_size/1.

### sparse_foldl(Function, InitialAcc::term(), Array::array()) -> term()

`Function = (Index::integer(), Value::term(), Acc::term()) -> term()`

Fold the elements of the array using the given function and
initial accumulator value, skipping default-valued entries. The
elements are visited in order from the lowest index to the highest.
If `Function`

is not a function, the call fails with reason `badarg`

.

*See also:* foldl/3, sparse_foldr/3.

### sparse_foldr(Function, InitialAcc::term(), Array::array()) -> term()

`Function = (Index::integer(), Value::term(), Acc::term()) -> term()`

Fold the elements of the array right-to-left using the given
function and initial accumulator value, skipping default-valued
entries. The elements are visited in order from the highest index to
the lowest. If `Function`

is not a function, the call fails with
reason `badarg`

.

*See also:* foldr/3, sparse_foldl/3.

### sparse_map(Function, Array::array()) -> array()

`Function = (Index::integer(), Value::term()) -> term()`

Map the given function onto each element of the array, skipping
default-valued entries. The elements are visited in order from the
lowest index to the highest. If `Function`

is not a function, the
call fails with reason `badarg`

.

*See also:* map/2.

### sparse_size(A::array()) -> integer()

### sparse_to_list(Array::array()) -> list()

### sparse_to_orddict(Array::array()) -> [{Index::integer(), Value::term()}]

Convert the array to an ordered list of pairs `{Index, Value}`

,
skipping default-valued entries.

*See also:* to_orddict/1.

### to_list(Array::array()) -> list()

### to_orddict(Array::array()) -> [{Index::integer(), Value::term()}]

Convert the array to an ordered list of pairs `{Index, Value}`

.

*See also:* from_orddict/2, sparse_to_orddict/1.