Some of the intrinsic functions have special properties:
- variadic
- polymorphic
- elemental
Variadic intrinsic functions are indicated with ....
Polymorphic
function parameters are indicated with special types. Rank-polymorphic array
function parameters are indicated with any_rank.
Elemental
functions are tagged with #elemental. Elemental functions can take either a
scalar argument or an array argument. When given an array argument, the result
is also an array, with the same rank and size as the argument, with values as if
the function had been called element-wise on each member of the array.
These are the special polymorphic types used in documentation:
any: any unlimitted primitive, built-in syntran typeany_num: any numeric type, i.e.i32,i64,f32, orf64any_float: any floating-point type, i.e.f32orf64
All of these special properties are special cases for intrinsic functions. For now at least, user-defined functions cannot be variadic, polymorphic, or elemental.
fn all(mask: [bool; any_rank])Return true if every element of mask is true, including the case where
mask is empty. Otherwise, i.e. if at least 1 element of mask is false,
return false.
Related functions: any
fn any(mask: [bool; any_rank])Return true if at least 1 element of mask is true. Otherwise, i.e. if
all elements of mask are false or mask is empty, return false.
Related functions: all
fn char(i: i32): strPerform an ASCII table lookup and return the single-character string with ASCII
index i. For example, char(65) is "A". The inverse of the ASCII lookup
is i32, which is also overloaded to do other casting operations.
Related functions: i32
fn close(file_handle: file)Close a file_handle created by open
Related functions: open, writeln
fn count(mask: [bool; any_rank]): i64Return how many elements of mask are true. If mask is empty, return 0.
Not to be confused with size.
fn eof(file_handle: file): boolHas the end of file (EOF) been reached yet while reading from file_handle? The return value is false until and including the last line is read and does not become true until reading past the EOF is attempted, for which readln() will return an empty string
Related functions: close, readln, open
fn exit(status: i32)Terminate the process normally,
return the system exit status status, and
perform the regular cleanup for terminating programs.
By the C convention, use status 0 for success and any other value for failure.
fn exp(x: any_float): any_floatCompute the base e exponential of x
fn i32(a: any_num): i32
#elementalExplicitly cast any numeric type to 32-bit integer i32. The i32() fn also
casts single-character strings to their ASCII values, e.g. i32("A") returns
65. The inverse ASCII lookup is char. To read a decimal number
from a string, use parse_i32() instead.
The i32 function is elemental, i.e. it can operate on both scalars and
arrays.
Other non-numeric types (bool, file) cannot be cast to i64.
Related functions: char
fn i64(a: any_num): i64
#elementalExplicitly cast any numeric type to 64-bit integer i64.
Other non-numeric types (bool, file) cannot be cast to i64.
fn len(s: str): i32Get the length of a string s
fn max(a0: any_num, a1: any_num, a2: any_num, ...): any_numReturn the argument with the largest (most positive) value. The arguments of
max must be homogeneous types, i.e. a1 must be the same type as a0 in one
invocation. For example, this is a parser error:
let x = max(1, 2.0);
// Error: function `max` parameter 1 `a1` requires type `i32` but was given `f64`
// --> <stdin>:1:8
// |
// 1 | max(1, 2.0);
// | ^^^ wrong argument typeSeperate invocations with different types are fine:
let max_int = max(1, 2);
let max_flt = max(3.0, 2.0);Related functions: min
fn min(a0: any_num, a1: any_num, a2: any_num, ...): any_numReturn the argument with the smallest (most negative) value. Like with max,
the arguments of min must be homogeneous types.
Related functions: max
fn open(filename: str): fileOpen a file handle named filename
Related functions: close, writeln
fn parse_f32(s: str): f32Convert from a string to a float f32. For example, parse_f32("65.4")
returns roughly 65.4f.
Any invalid numbers will cause a runtime error.
Note that other numeric types (i32) can be converted to float simply by
multiplying by 1.0f.
Related functions: parse_i32, str
fn parse_f64(s: str): f64Convert from a string to a double f64. For example, parse_f64("65.4")
returns roughly 65.4.
Any invalid numbers will cause a runtime error.
Note that other numeric types (i32) can be converted to double simply by
multiplying by 1.0.
Related functions: parse_f32, str
fn parse_i32(s: str): i32Convert from a string to an integer i32. For example, parse_i32("65")
returns 65. This is in contrast to calling i32() on character arguments,
which performs an ASCII table lookup.
Any invalid numbers will cause a runtime error. Other languages variously refer
to similar functions as atoi(), stoi(), read(), parse(), strtol(),
str2int(), etc.
Note that other numeric types (f32) can be implicitly cast to i32.
Related functions: parse_i64, str
fn parse_i64(s: str): i64Convert from a string to an integer i64. For example, parse_i64("65")
returns 65.
Any invalid numbers will cause a runtime error.
Related functions: parse_i32, str
fn println(s0: any, s1: any, s2: any, ...)Print to the standard output, with a newline
Related functions: str, writeln
fn readln(file_handle: file): strRead a single line from file_handle
Related functions: close, eof, open
fn size(array: [any; any_rank], dim: i32): i32Determine the extent of array along a specified dimension dim
fn str(s0: any, s1: any, s2: any, ...): strConvert arguments to str and concatenate
Related functions: println, writeln, parse_i32
fn sum(array: [any_num; any_rank]): any_numReturn the sum of all elements in an array.
fn writeln(file_handle: file, s0: any, s1: any, s2: any, ...)Write to file_handle, with a newline