BREAKING: Migrate math functions to keyword args (#6491)

This commit is contained in:
Jonathan Tran
2025-04-26 19:33:41 -04:00
committed by GitHub
parent d7e80b3cc7
commit 0f88598dc0
32 changed files with 586 additions and 537 deletions

View File

@ -4,9 +4,9 @@ use anyhow::Result;
use kcl_derive_docs::stdlib;
use crate::{
errors::{KclError, KclErrorDetails},
errors::KclError,
execution::{
types::{NumericType, RuntimeType, UnitAngle, UnitType},
types::{ArrayLen, NumericType, RuntimeType, UnitAngle, UnitType},
ExecState, KclValue,
},
std::args::{Args, TyF64},
@ -147,8 +147,8 @@ fn inner_pi() -> Result<f64, KclError> {
/// Compute the square root of a number.
pub async fn sqrt(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
let result = inner_sqrt(num.n)?;
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::num_any(), exec_state)?;
let result = inner_sqrt(input.n);
Ok(args.make_user_val_from_f64_with_type(TyF64::new(result, exec_state.current_default_units())))
}
@ -170,17 +170,22 @@ pub async fn sqrt(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kc
#[stdlib {
name = "sqrt",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to compute the square root of."},
}
}]
fn inner_sqrt(num: f64) -> Result<f64, KclError> {
Ok(num.sqrt())
fn inner_sqrt(input: f64) -> f64 {
input.sqrt()
}
/// Compute the absolute value of a number.
pub async fn abs(_exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
let result = inner_abs(num.n)?;
pub async fn abs(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::num_any(), exec_state)?;
let result = inner_abs(input.n);
Ok(args.make_user_val_from_f64_with_type(num.map_value(result)))
Ok(args.make_user_val_from_f64_with_type(input.map_value(result)))
}
/// Compute the absolute value of a number.
@ -207,17 +212,22 @@ pub async fn abs(_exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kc
#[stdlib {
name = "abs",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to compute the absolute value of."},
}
}]
fn inner_abs(num: f64) -> Result<f64, KclError> {
Ok(num.abs())
fn inner_abs(input: f64) -> f64 {
input.abs()
}
/// Round a number to the nearest integer.
pub async fn round(_exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
let result = inner_round(num.n)?;
pub async fn round(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::num_any(), exec_state)?;
let result = inner_round(input.n);
Ok(args.make_user_val_from_f64_with_type(num.map_value(result)))
Ok(args.make_user_val_from_f64_with_type(input.map_value(result)))
}
/// Round a number to the nearest integer.
@ -235,17 +245,22 @@ pub async fn round(_exec_state: &mut ExecState, args: Args) -> Result<KclValue,
#[stdlib {
name = "round",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to round."},
}
}]
fn inner_round(num: f64) -> Result<f64, KclError> {
Ok(num.round())
fn inner_round(input: f64) -> f64 {
input.round()
}
/// Compute the largest integer less than or equal to a number.
pub async fn floor(_exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
let result = inner_floor(num.n)?;
pub async fn floor(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::num_any(), exec_state)?;
let result = inner_floor(input.n);
Ok(args.make_user_val_from_f64_with_type(num.map_value(result)))
Ok(args.make_user_val_from_f64_with_type(input.map_value(result)))
}
/// Compute the largest integer less than or equal to a number.
@ -263,17 +278,22 @@ pub async fn floor(_exec_state: &mut ExecState, args: Args) -> Result<KclValue,
#[stdlib {
name = "floor",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to round."},
}
}]
fn inner_floor(num: f64) -> Result<f64, KclError> {
Ok(num.floor())
fn inner_floor(input: f64) -> f64 {
input.floor()
}
/// Compute the smallest integer greater than or equal to a number.
pub async fn ceil(_exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
let result = inner_ceil(num.n)?;
pub async fn ceil(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::num_any(), exec_state)?;
let result = inner_ceil(input.n);
Ok(args.make_user_val_from_f64_with_type(num.map_value(result)))
Ok(args.make_user_val_from_f64_with_type(input.map_value(result)))
}
/// Compute the smallest integer greater than or equal to a number.
@ -291,14 +311,23 @@ pub async fn ceil(_exec_state: &mut ExecState, args: Args) -> Result<KclValue, K
#[stdlib {
name = "ceil",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to round."},
}
}]
fn inner_ceil(num: f64) -> Result<f64, KclError> {
Ok(num.ceil())
fn inner_ceil(input: f64) -> f64 {
input.ceil()
}
/// Compute the minimum of the given arguments.
pub async fn min(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let nums = args.get_number_array_with_types()?;
let nums: Vec<TyF64> = args.get_unlabeled_kw_arg_typed(
"input",
&RuntimeType::Array(Box::new(RuntimeType::num_any()), ArrayLen::None),
exec_state,
)?;
let (nums, ty) = NumericType::combine_eq_array(&nums);
if ty == NumericType::Unknown {
exec_state.warn(CompilationError::err(
@ -318,7 +347,7 @@ pub async fn min(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kcl
/// |> startProfile(at = [0, 0])
/// |> angledLine(
/// angle = 70,
/// length = min(15, 31, 4, 13, 22)
/// length = min([15, 31, 4, 13, 22])
/// )
/// |> line(end = [20, 0])
/// |> close()
@ -328,12 +357,17 @@ pub async fn min(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kcl
#[stdlib {
name = "min",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "An array of numbers to compute the minimum of."},
}
}]
fn inner_min(args: Vec<f64>) -> f64 {
fn inner_min(input: Vec<f64>) -> f64 {
let mut min = f64::MAX;
for arg in args.iter() {
if *arg < min {
min = *arg;
for num in input.iter() {
if *num < min {
min = *num;
}
}
@ -342,7 +376,11 @@ fn inner_min(args: Vec<f64>) -> f64 {
/// Compute the maximum of the given arguments.
pub async fn max(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let nums = args.get_number_array_with_types()?;
let nums: Vec<TyF64> = args.get_unlabeled_kw_arg_typed(
"input",
&RuntimeType::Array(Box::new(RuntimeType::num_any()), ArrayLen::None),
exec_state,
)?;
let (nums, ty) = NumericType::combine_eq_array(&nums);
if ty == NumericType::Unknown {
exec_state.warn(CompilationError::err(
@ -362,7 +400,7 @@ pub async fn max(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kcl
/// |> startProfile(at = [0, 0])
/// |> angledLine(
/// angle = 70,
/// length = max(15, 31, 4, 13, 22)
/// length = max([15, 31, 4, 13, 22])
/// )
/// |> line(end = [20, 0])
/// |> close()
@ -372,12 +410,17 @@ pub async fn max(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kcl
#[stdlib {
name = "max",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "An array of numbers to compute the maximum of."},
}
}]
fn inner_max(args: Vec<f64>) -> f64 {
fn inner_max(input: Vec<f64>) -> f64 {
let mut max = f64::MIN;
for arg in args.iter() {
if *arg > max {
max = *arg;
for num in input.iter() {
if *num > max {
max = *num;
}
}
@ -386,22 +429,9 @@ fn inner_max(args: Vec<f64>) -> f64 {
/// Compute the number to a power.
pub async fn pow(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let nums = args.get_number_array_with_types()?;
if nums.len() > 2 {
return Err(KclError::Type(KclErrorDetails {
message: format!("expected 2 arguments, got {}", nums.len()),
source_ranges: vec![args.source_range],
}));
}
if nums.len() <= 1 {
return Err(KclError::Type(KclErrorDetails {
message: format!("expected 2 arguments, got {}", nums.len()),
source_ranges: vec![args.source_range],
}));
}
let result = inner_pow(nums[0].n, nums[1].n)?;
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::num_any(), exec_state)?;
let exp: TyF64 = args.get_kw_arg_typed("exp", &RuntimeType::count(), exec_state)?;
let result = inner_pow(input.n, exp.n);
Ok(args.make_user_val_from_f64_with_type(TyF64::new(result, exec_state.current_default_units())))
}
@ -413,7 +443,7 @@ pub async fn pow(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kcl
/// |> startProfile(at = [0, 0])
/// |> angledLine(
/// angle = 50,
/// length = pow(5, 2),
/// length = pow(5, exp = 2),
/// )
/// |> yLine(endAbsolute = 0)
/// |> close()
@ -423,27 +453,21 @@ pub async fn pow(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kcl
#[stdlib {
name = "pow",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to raise."},
exp = {docs = "The power to raise to."},
}
}]
fn inner_pow(num: f64, pow: f64) -> Result<f64, KclError> {
Ok(num.powf(pow))
fn inner_pow(input: f64, exp: f64) -> f64 {
input.powf(exp)
}
/// Compute the arccosine of a number (in radians).
pub async fn acos(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
if matches!(
num.ty,
NumericType::Default {
angle: UnitAngle::Degrees,
..
}
) {
exec_state.warn(CompilationError::err(
args.source_range,
"`acos` requires its input in radians, but the input is assumed to be in degrees. You can use a numeric suffix (e.g., `0rad`) or type ascription (e.g., `(1/2): number(rad)`) to show the number is in radians, or `toRadians` to convert from degrees to radians",
));
}
let result = inner_acos(num.n)?;
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::count(), exec_state)?;
let result = inner_acos(input.n);
Ok(args.make_user_val_from_f64_with_type(TyF64::new(result, NumericType::radians())))
}
@ -466,27 +490,20 @@ pub async fn acos(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kc
#[stdlib {
name = "acos",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to compute arccosine of."},
}
}]
fn inner_acos(num: f64) -> Result<f64, KclError> {
Ok(num.acos())
fn inner_acos(input: f64) -> f64 {
input.acos()
}
/// Compute the arcsine of a number (in radians).
pub async fn asin(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
if matches!(
num.ty,
NumericType::Default {
angle: UnitAngle::Degrees,
..
}
) {
exec_state.warn(CompilationError::err(
args.source_range,
"`asin` requires its input in radians, but the input is assumed to be in degrees. You can use a numeric suffix (e.g., `0rad`) or type ascription (e.g., `(1/2): number(rad)`) to show the number is in radians, or `toRadians` to convert from degrees to radians",
));
}
let result = inner_asin(num.n)?;
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::count(), exec_state)?;
let result = inner_asin(input.n);
Ok(args.make_user_val_from_f64_with_type(TyF64::new(result, NumericType::radians())))
}
@ -508,33 +525,28 @@ pub async fn asin(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kc
#[stdlib {
name = "asin",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to compute arcsine of."},
}
}]
fn inner_asin(num: f64) -> Result<f64, KclError> {
Ok(num.asin())
fn inner_asin(input: f64) -> f64 {
input.asin()
}
/// Compute the arctangent of a number (in radians).
pub async fn atan(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
if matches!(
num.ty,
NumericType::Default {
angle: UnitAngle::Degrees,
..
}
) {
exec_state.warn(CompilationError::err(
args.source_range,
"`atan` requires its input in radians, but the input is assumed to be in degrees. You can use a numeric suffix (e.g., `0rad`) or type ascription (e.g., `(1/2): number(rad)`) to show the number is in radians, or `toRadians` to convert from degrees to radians",
));
}
let result = inner_atan(num.n)?;
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::count(), exec_state)?;
let result = inner_atan(input.n);
Ok(args.make_user_val_from_f64_with_type(TyF64::new(result, NumericType::radians())))
}
/// Compute the arctangent of a number (in radians).
///
/// Consider using `atan2()` instead for the true inverse of tangent.
///
/// ```no_run
/// sketch001 = startSketchOn('XZ')
/// |> startProfile(at = [0, 0])
@ -550,9 +562,14 @@ pub async fn atan(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kc
#[stdlib {
name = "atan",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to compute arctangent of."},
}
}]
fn inner_atan(num: f64) -> Result<f64, KclError> {
Ok(num.atan())
fn inner_atan(input: f64) -> f64 {
input.atan()
}
/// Compute the four quadrant arctangent of Y and X (in radians).
@ -560,7 +577,7 @@ pub async fn atan2(exec_state: &mut ExecState, args: Args) -> Result<KclValue, K
let y = args.get_kw_arg_typed("y", &RuntimeType::length(), exec_state)?;
let x = args.get_kw_arg_typed("x", &RuntimeType::length(), exec_state)?;
let (y, x, _) = NumericType::combine_eq_coerce(y, x);
let result = inner_atan2(y, x)?;
let result = inner_atan2(y, x);
Ok(args.make_user_val_from_f64_with_type(TyF64::new(result, NumericType::radians())))
}
@ -589,8 +606,8 @@ pub async fn atan2(exec_state: &mut ExecState, args: Args) -> Result<KclValue, K
x = { docs = "X"},
}
}]
fn inner_atan2(y: f64, x: f64) -> Result<f64, KclError> {
Ok(y.atan2(x))
fn inner_atan2(y: f64, x: f64) -> f64 {
y.atan2(x)
}
/// Compute the logarithm of the number with respect to an arbitrary base.
@ -599,21 +616,9 @@ fn inner_atan2(y: f64, x: f64) -> Result<f64, KclError> {
/// details; `log2()` can produce more accurate results for base 2,
/// and `log10()` can produce more accurate results for base 10.
pub async fn log(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let nums = args.get_number_array_with_types()?;
if nums.len() > 2 {
return Err(KclError::Type(KclErrorDetails {
message: format!("expected 2 arguments, got {}", nums.len()),
source_ranges: vec![args.source_range],
}));
}
if nums.len() <= 1 {
return Err(KclError::Type(KclErrorDetails {
message: format!("expected 2 arguments, got {}", nums.len()),
source_ranges: vec![args.source_range],
}));
}
let result = inner_log(nums[0].n, nums[1].n)?;
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::num_any(), exec_state)?;
let base: TyF64 = args.get_kw_arg_typed("base", &RuntimeType::count(), exec_state)?;
let result = inner_log(input.n, base.n);
Ok(args.make_user_val_from_f64_with_type(TyF64::new(result, exec_state.current_default_units())))
}
@ -627,7 +632,7 @@ pub async fn log(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kcl
/// ```no_run
/// exampleSketch = startSketchOn("XZ")
/// |> startProfile(at = [0, 0])
/// |> line(end = [log(100, 5), 0])
/// |> line(end = [log(100, base = 5), 0])
/// |> line(end = [5, 8])
/// |> line(end = [-10, 0])
/// |> close()
@ -637,15 +642,21 @@ pub async fn log(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kcl
#[stdlib {
name = "log",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to compute the logarithm of."},
base = {docs = "The base of the logarithm."},
}
}]
fn inner_log(num: f64, base: f64) -> Result<f64, KclError> {
Ok(num.log(base))
fn inner_log(input: f64, base: f64) -> f64 {
input.log(base)
}
/// Compute the base 2 logarithm of the number.
pub async fn log2(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
let result = inner_log2(num.n)?;
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::num_any(), exec_state)?;
let result = inner_log2(input.n);
Ok(args.make_user_val_from_f64_with_type(TyF64::new(result, exec_state.current_default_units())))
}
@ -665,15 +676,20 @@ pub async fn log2(exec_state: &mut ExecState, args: Args) -> Result<KclValue, Kc
#[stdlib {
name = "log2",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to compute the logarithm of."},
}
}]
fn inner_log2(num: f64) -> Result<f64, KclError> {
Ok(num.log2())
fn inner_log2(input: f64) -> f64 {
input.log2()
}
/// Compute the base 10 logarithm of the number.
pub async fn log10(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
let result = inner_log10(num.n)?;
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::num_any(), exec_state)?;
let result = inner_log10(input.n);
Ok(args.make_user_val_from_f64_with_type(TyF64::new(result, exec_state.current_default_units())))
}
@ -694,14 +710,14 @@ pub async fn log10(exec_state: &mut ExecState, args: Args) -> Result<KclValue, K
name = "log10",
tags = ["math"],
}]
fn inner_log10(num: f64) -> Result<f64, KclError> {
Ok(num.log10())
fn inner_log10(num: f64) -> f64 {
num.log10()
}
/// Compute the natural logarithm of the number.
pub async fn ln(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclError> {
let num = args.get_number_with_type()?;
let result = inner_ln(num.n)?;
let input: TyF64 = args.get_unlabeled_kw_arg_typed("input", &RuntimeType::num_any(), exec_state)?;
let result = inner_ln(input.n);
Ok(args.make_user_val_from_f64_with_type(TyF64::new(result, exec_state.current_default_units())))
}
@ -721,9 +737,14 @@ pub async fn ln(exec_state: &mut ExecState, args: Args) -> Result<KclValue, KclE
#[stdlib {
name = "ln",
tags = ["math"],
keywords = true,
unlabeled_first = true,
args = {
input = {docs = "The number to compute the logarithm of."},
}
}]
fn inner_ln(num: f64) -> Result<f64, KclError> {
Ok(num.ln())
fn inner_ln(input: f64) -> f64 {
input.ln()
}
/// Return the value of Eulers number `e`.