7.3 KiB
id | title |
---|---|
loops | Loops |
General Iteration
General iteration in PascaLIGO takes the shape of general loops, which
should be familiar to programmers of imperative languages as "while
loops". Those loops are of the form while <condition> <block>
. Their
associated block is repeatedly evaluated until the condition becomes
true, or never evaluated if the condition is false at the start. The
loop never terminates if the condition never becomes true. Because we
are writing smart contracts on Tezos, when the condition of a "while"
loops fails to become true, the execution will run out of gas and stop
with a failure anyway.
Here is how to compute the greatest common divisors of two natural number by means of Euclid's algorithm:
function gcd (var x : nat; var y : nat) : nat is
block {
if x < y then
block {
const z : nat = x;
x := y; y := z
}
else skip;
var r : nat := 0n;
while y =/= 0n block {
r := x mod y;
x := y;
y := r
}
} with x
You can call the function gcd
defined above using the LIGO compiler
like so:
ligo run-function
gitlab-pages/docs/language-basics/src/loops/gcd.ligo gcd '(2n*2n*3n*11n, 2n*2n*2n*3n*3n*5n*7n)'
# Outputs: +12
CameLIGO is a functional language where user-defined values are constant, therefore it makes no sense in CameLIGO to feature loops, which we understand as syntactic constructs where the state of a stopping condition is mutated, as with "while" loops in PascaLIGO.
Instead, CameLIGO features a fold operation as a predefined function
named Loop.fold_while
. It takes an initial value of a certain type,
called an accumulator, and repeatedly calls a given function, called
iterated function, that takes that accumulator and returns the next
value of the accumulator, until a condition is met and the fold stops
with the final value of the accumulator. The iterated function needs
to have a special type: if the type of the accumulator is t
, then it
must have the type bool * t
(not simply t
). It is the boolean
value that denotes whether the stopping condition has been reached.
Here is how to compute the greatest common divisors of two natural number by means of Euclid's algorithm:
let iter (x,y : nat * nat) : bool * (nat * nat) =
if y = 0n then false, (x,y) else true, (y, x mod y)
let gcd (x,y : nat * nat) : nat =
let x,y = if x < y then y,x else x,y in
let x,y = Loop.fold_while iter (x,y)
in x
To ease the writing and reading of the iterated functions (here,
iter
), two predefined functions are provided: continue
and stop
:
let iter (x,y : nat * nat) : bool * (nat * nat) =
if y = 0n then stop (x,y) else continue (y, x mod y)
let gcd (x,y : nat * nat) : nat =
let x,y = if x < y then y,x else x,y in
let x,y = Loop.fold_while iter (x,y)
in x
You can call the function gcd
defined above using the LIGO compiler
like so:
ligo run-function
gitlab-pages/docs/language-basics/src/loops/gcd.mligo gcd (2n*2n*3n*11n, 2n*2n*2n*3n*3n*5n*7n)'
# Outputs: +12
ReasonLIGO is a functional language where user-defined values are constant, therefore it makes no sense in ReasonLIGO to feature loops, which we understand as syntactic constructs where the state of a stopping condition is mutated, as with "while" loops in PascaLIGO.
Instead, ReasonLIGO features a fold operation as a predefined
function named Loop.fold_while
. It takes an initial value of a
certain type, called an accumulator, and repeatedly calls a given
function, called iterated function, that takes that accumulator and
returns the next value of the accumulator, until a condition is met
and the fold stops with the final value of the accumulator. The
iterated function needs to have a special type: if the type of the
accumulator is t
, then it must have the type bool * t
(not simply
t
). It is the boolean value that denotes whether the stopping
condition has been reached.
Here is how to compute the greatest common divisors of two natural number by means of Euclid's algorithm:
let iter = ((x,y) : (nat, nat)) : (bool, (nat, nat)) =>
if (y == 0n) { (false, (x,y)); } else { (true, (y, x mod y)); };
let gcd = ((x,y) : (nat, nat)) : nat => {
let (x,y) = if (x < y) { (y,x); } else { (x,y); };
let (x,y) = Loop.fold_while (iter, (x,y));
x
};
To ease the writing and reading of the iterated functions (here,
iter
), two predefined functions are provided: continue
and stop
:
let iter = ((x,y) : (nat, nat)) : (bool, (nat, nat)) =>
if (y == 0n) { stop ((x,y)); } else { continue ((y, x mod y)); };
let gcd = ((x,y) : (nat, nat)) : nat => {
let (x,y) = if (x < y) { (y,x); } else { (x,y); };
let (x,y) = Loop.fold_while (iter, (x,y));
x
};
Bounded Loops
In addition to general loops, PascaLIGO features a specialised kind of
loop to iterate over bounded intervals. These loops are familiarly
known as "for loops" and they have the form for <variable assignment> to <upper bound> <block>
, which is familiar for programmers of
imperative languages.
Consider how to sum integers from 0
to n
:
function sum (var n : nat) : int is block {
var acc : int := 0;
for i := 1 to int (n) block {
acc := acc + i
}
} with acc
(Please do not use that function: there exists a closed form formula.)
You can call the function sum
defined above using the LIGO compiler
like so:
ligo run-function
gitlab-pages/docs/language-basics/src/loops/sum.ligo sum 7n
# Outputs: 28
PascaLIGO "for" loops can also iterate through the contents of a
collection, that is, a list, a set or a map. This is done with a loop
of the form for <element var> in <collection type> <collection var> <block>
, where <collection type
is any of the following keywords:
list
, set
or map
.
Here is an example where the integers in a list are summed up.
function sum_list (var l : list (int)) : int is block {
var total : int := 0;
for i in list l block {
total := total + i
}
} with total
You can call the function sum_list
defined above using the LIGO compiler
like so:
ligo run-function
gitlab-pages/docs/language-basics/src/loops/collection.ligo sum_list
'list [1;2;3]'
# Outputs: 6
Here is an example where the integers in a set are summed up.
function sum_set (var s : set (int)) : int is block {
var total : int := 0;
for i in set s block {
total := total + i
}
} with total
You can call the function sum_set
defined above using the LIGO compiler
like so:
ligo run-function
gitlab-pages/docs/language-basics/src/loops/collection.ligo sum_set
'set [1;2;3]'
# Outputs: 6
Loops over maps are actually loops over the bindings of the map, that
is, a pair key-value noted key -> value
(or any other
variables). Give a map from strings to integers, here is how to sum
all the integers and concatenate all the strings.
You can call the function sum_map
defined above using the LIGO compiler
like so:
ligo run-function
gitlab-pages/docs/language-basics/src/loops/collection.ligo sum_map
'map ["1"->1; "2"->2; "3"->3]'
# Outputs: ( "123", 6 )