Dynamic arrays.
A dynamic array is equivalent to a OCaml array that will resize itself when elements are added or removed, except that floats are boxed and that no initialization element is required.
When an operation on an array fails, Invalid_arg
is raised. The
integer is the value that made the operation fail, the first string
contains the function name that has been called and the second string
contains the parameter name that made the operation fail.
make count
returns an array with some memory already allocated so
up to count
elements can be stored into it without resizing.
init n f
returns an array of n
elements filled with values
returned by f 0 , f 1, ... f (n-1)
.
get darr idx
gets the element in darr
at index idx
. If darr
has
len
elements in it, then the valid indexes range from 0
to len-1
.
set darr idx v
sets the element of darr
at index idx
to value
v
. The previous value is overwritten.
insert darr idx v
inserts v
into darr
at index idx
. All elements
of darr
with an index greater than or equal to idx
have their
index incremented (are moved up one place) to make room for the new
element.
add darr v
appends v
onto darr
. v
becomes the new
last element of darr
.
delete darr idx
deletes the element of darr
at idx
. All elements
with an index greater than idx
have their index decremented (are
moved down one place) to fill in the hole.
delete_last darr
deletes the last element of darr
. This is equivalent
of doing delete darr ((length darr) - 1)
.
delete_range darr p len
deletes len
elements starting at index p
.
All elements with an index greater than p+len
are moved to fill
in the hole.
of_list lst
returns a dynamic array with the elements of lst
in
it in order.
of_array arr
returns an array with the elements of arr
in it
in order.
iter f darr
calls the function f
on every element of darr
. It
is equivalent to for i = 0 to length darr - 1 do f (get darr i) done;
iteri f darr
calls the function f
on every element of darr
. It
is equivalent to for i = 0 to length darr - 1 do f i (get darr i) done;
fold_left f x darr
computes
f ( ... ( f ( f (get darr 0) x) (get darr 1) ) ... ) (get darr n-1)
,
similar to Array.fold_left
or List.fold_left
.
fold_right f darr x
computes
f (get darr 0) (f (get darr 1) ( ... ( f (get darr n-1) x ) ... ) )
similar to Array.fold_right
or List.fold_right
.
index_of f darr
returns the index of the first element x
in darr such
as f x
returns true
or raise Not_found
if not found.
The type of a resizer function.
Resizer functions are called whenever elements are added to or removed from the dynamic array to determine what the current number of storage spaces in the array should be. The three named arguments passed to a resizer are the current number of storage spaces in the array, the length of the array before the elements are added or removed, and the length the array will be after the elements are added or removed. If elements are being added, newlength will be larger than oldlength, if elements are being removed, newlength will be smaller than oldlength. If the resizer function returns exactly oldlength, the size of the array is only changed when adding an element while there is not enough space for it.
By default, all dynamic arrays are created with the default_resizer
.
When a dynamic array is created from another dynamic array (using copy
,
map
, etc. ) the resizer of the copy will be the same as the original
dynamic array resizer. To change the resizer, use the set_resizer
function.
The default resizer function the library is using - in this version
of DynArray, this is the exponential_resizer
but should change in
next versions.
The exponential resizer- The default resizer except when the resizer is being copied from some other darray.
exponential_resizer
works by doubling or halving the number of
slots until they "fit". If the number of slots is less than the
new length, the number of slots is doubled until it is greater
than the new length (or Sys.max_array_size is reached).
If the number of slots is more than four times the new length, the number of slots is halved until it is less than four times the new length.
Allowing darrays to fall below 25% utilization before shrinking them prevents "thrashing". Consider the case where the caller is constantly adding a few elements, and then removing a few elements, causing the length to constantly cross above and below a power of two. Shrinking the array when it falls below 50% would causing the underlying array to be constantly allocated and deallocated. A few elements would be added, causing the array to be reallocated and have a usage of just above 50%. Then a few elements would be remove, and the array would fall below 50% utilization and be reallocated yet again. The bulk of the array, untouched, would be copied and copied again. By setting the threshold at 25% instead, such "thrashing" only occurs with wild swings- adding and removing huge numbers of elements (more than half of the elements in the array).
exponential_resizer
is a good performing resizer for most
applications. A list allocates 2 words for every element, while an
array (with large numbers of elements) allocates only 1 word per
element (ignoring unboxed floats). On insert, exponential_resizer
keeps the amount of wasted "extra" array elements below 50%, meaning
that less than 2 words per element are used. Even on removals
where the amount of wasted space is allowed to rise to 75%, that
only means that darray is using 4 words per element. This is
generally not a significant overhead.
Furthermore, exponential_resizer
minimizes the number of copies
needed- appending n elements into an empty darray with initial size
0 requires between n and 2n elements of the array be copied- O(n)
work, or O(1) work per element (on average). A similar argument
can be made that deletes from the end of the array are O(1) as
well (obviously deletes from anywhere else are O(n) work- you
have to move the n or so elements above the deleted element down).
The stepwise resizer- another example of a resizer function, this time of a parameterized resizer.
The resizer returned by step_resizer step
returns the smallest
multiple of step
larger than newlength
if currslots
is less
then newlength
-step
or greater than newlength
.
For example, to make an darray with a step of 10, a length
of len, and a null of null, you would do:
make
~resizer:(step_resizer
10) len null
conservative_exponential_resizer
is an example resizer function
which uses the oldlength parameter. It only shrinks the array
on inserts- no deletes shrink the array, only inserts. It does
this by comparing the oldlength and newlength parameters. Other
than that, it acts like exponential_resizer
.
*