9-fs / scaler

scaler

Introduction

This crate provides a convenient Formatter to scale, round, and display numbers.

Scaling describes the usage of decimal or binary unit prefixes to increase readability; though no scaling and scientific notation are also supported.
Rounding can be done either to a specified magnitude or to a number of significant digits.
Separators can be freely adjusted. The group separator separates groups of digits every 3 digits before the decimal separator, while the decimal separator separates the integer and fractional parts of a number.
The sign behaviour can be set to always show the sign or only show the sign when the number is negative.

Installation

The feature warn_about_problematic_separators warns using log::warn! if separators are being set with Formatter::set_separators that could lead to ambiguous formatting. It depends on the log crate and is the only dependency. If a dependencyless build should be desired, it can be disabled by specifying default-features = false in your Cargo.toml entry.

Usage

1. Execute Formatter::new to create a new Formatter with default settings.
2. Adjust separators, rounding, scaling, and sign behaviour as necessary using the setters.
3. Format numbers with Formatter::format.

Rounding

Examples have scaling disabled for easier understanding.

• Magnitude:

  • Round to digit at magnitude $10^m$.
  • Contains $m$.

  let f: scaler::Formatter = scaler::Formatter::new()
      .set_scaling(scaler::Scaling::None)
      .set_rounding(scaler::Rounding::Magnitude(-2));
  assert_eq!(f.format(123.456), "123,46");
  assert_eq!(f.format(0.789), "0,79");
  assert_eq!(f.format(42069), "42.069,00");

  let f: scaler::Formatter = scaler::Formatter::new()
      .set_scaling(scaler::Scaling::None)
      .set_rounding(scaler::Rounding::Magnitude(-1));
  assert_eq!(f.format(123.456), "123,5");
  assert_eq!(f.format(0.789), "0,8");
  assert_eq!(f.format(42069), "42.069,0");

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::None)
     .set_rounding(scaler::Rounding::Magnitude(0));
  assert_eq!(f.format(123.456), "123");
  assert_eq!(f.format(0.789), "1");
  assert_eq!(f.format(42069), "42.069");

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::None)
     .set_rounding(scaler::Rounding::Magnitude(1));
  assert_eq!(f.format(123.456), "120");
  assert_eq!(f.format(0.789), "0");
  assert_eq!(f.format(42069), "42.070");

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::None)
     .set_rounding(scaler::Rounding::Magnitude(2));
  assert_eq!(f.format(123.456), "100");
  assert_eq!(f.format(0.789), "0");
  assert_eq!(f.format(42069), "42.100");

• SignificantDigits:

  • Round to $n$ significant numbers.
  • Contains $n$.

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::None)
     .set_rounding(scaler::Rounding::SignificantDigits(0));
  assert_eq!(f.format(123.456), "0");
  assert_eq!(f.format(0.789), "0");
  assert_eq!(f.format(42069), "0");

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::None)
     .set_rounding(scaler::Rounding::SignificantDigits(1));
  assert_eq!(f.format(123.456), "100");
  assert_eq!(f.format(0.789), "0,8");
  assert_eq!(f.format(42069), "40.000");

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::None)
     .set_rounding(scaler::Rounding::SignificantDigits(2));
  assert_eq!(f.format(123.456), "120");
  assert_eq!(f.format(0.789), "0,79");
  assert_eq!(f.format(42069), "42.000");

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::None)
     .set_rounding(scaler::Rounding::SignificantDigits(3));
  assert_eq!(f.format(123.456), "123");
  assert_eq!(f.format(0.789), "0,789");
  assert_eq!(f.format(42069), "42.100");

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::None)
     .set_rounding(scaler::Rounding::SignificantDigits(4));
  assert_eq!(f.format(123.456), "123,5");
  assert_eq!(f.format(0.789), "0,7890");
  assert_eq!(f.format(42069), "42.070");

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::None)
     .set_rounding(scaler::Rounding::SignificantDigits(5));
  assert_eq!(f.format(123.456), "123,46");
  assert_eq!(f.format(0.789), "0,78900");
  assert_eq!(f.format(42069), "42.069");

Scaling

• Binary:

  • Scales by factor $2^(10) = 1024$.
  • If no prefix for that magnitude defined: Fallback to scientific notation.
  • Contains whether or not to put space between number and unit prefix.

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::Binary(true));
  assert_eq!(f.format(0.5), "1,000 * 2^(-1)");
  assert_eq!(f.format(1), "1,000");
  assert_eq!(f.format(64), "64,00");
  assert_eq!(f.format(128), "128,0");
  assert_eq!(f.format(1023), "1.023");
  assert_eq!(f.format(1024), "1,000 Ki");
  assert_eq!(f.format(2_f64.powi(10)), "1,000 Ki");
  assert_eq!(f.format(2_f64.powi(20)), "1,000 Mi");
  assert_eq!(f.format(2_f64.powi(30)), "1,000 Gi");
  assert_eq!(f.format(2_f64.powi(40)), "1,000 Ti");
  assert_eq!(f.format(2_f64.powi(50)), "1,000 Pi");
  assert_eq!(f.format(2_f64.powi(60)), "1,000 Ei");
  assert_eq!(f.format(2_f64.powi(70)), "1,000 Zi");
  assert_eq!(f.format(2_f64.powi(80)), "1,000 Yi");
  assert_eq!(f.format(2_f64.powi(90)), "1,000 * 2^(90)");

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::Binary(false));
  assert_eq!(f.format(1024), "1,000Ki");

• Decimal:

  • Scales by factor $10^(3) = 1000$.
  • If no prefix for that magnitude defined: Fallback to scientific notation.
  • Contains whether or not to put space between number and unit prefix.

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::Decimal(true));
  assert_eq!(f.format(1e-31), "1,000 * 10^(-31)");
  assert_eq!(f.format(1e-30), "1,000 q");
  assert_eq!(f.format(1e-27), "1,000 r");
  assert_eq!(f.format(1e-24), "1,000 y");
  assert_eq!(f.format(1e-21), "1,000 z");
  assert_eq!(f.format(1e-18), "1,000 a");
  assert_eq!(f.format(1e-15), "1,000 f");
  assert_eq!(f.format(1e-12), "1,000 p");
  assert_eq!(f.format(1e-9), "1,000 n");
  assert_eq!(f.format(1e-6), "1,000 µ");
  assert_eq!(f.format(1e-3), "1,000 m");
  assert_eq!(f.format(1), "1,000");
  assert_eq!(f.format(10), "10,00");
  assert_eq!(f.format(100), "100,0");
  assert_eq!(f.format(999), "999,0");
  assert_eq!(f.format(1000), "1,000 k");
  assert_eq!(f.format(1e3), "1,000 k");
  assert_eq!(f.format(1e6), "1,000 M");
  assert_eq!(f.format(1e9), "1,000 G");
  assert_eq!(f.format(1e12), "1,000 T");
  assert_eq!(f.format(1e15), "1,000 P");
  assert_eq!(f.format(1e18), "1,000 E");
  assert_eq!(f.format(1e21), "1,000 Z");
  assert_eq!(f.format(1e24), "1,000 Y");
  assert_eq!(f.format(1e27), "1,000 R");
  assert_eq!(f.format(1e30), "1,000 Q");
  assert_eq!(f.format(1e33), "1,000 * 10^(33)");

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::Decimal(false));
  assert_eq!(f.format(1000), "1,000k");

• None:

  • no scaling
  • no fallback to scientific notation

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::None);
  assert_eq!(f.format(1e-10), "0,0000000001000");
  assert_eq!(f.format(0.1), "0,1000");
  assert_eq!(f.format(1), "1,000");
  assert_eq!(f.format(10), "10,00");
  assert_eq!(f.format(100), "100,0");
  assert_eq!(f.format(1000), "1.000");
  assert_eq!(f.format(1e10), "10.000.000.000");

• Scientific:

  • always scientific notation

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_scaling(scaler::Scaling::Scientific);
  assert_eq!(f.format(0.1), "1,000 * 10^(-1)");
  assert_eq!(f.format(1), "1,000 * 10^(0)");
  assert_eq!(f.format(10), "1,000 * 10^(1)");

Separators

• group_separator
  • Separates groups every 3 digits before the decimal separator.
• decimal_separator
  • Separates the integer and fractional parts of a number.

Examples have scaling disabled for easier understanding.

let f: scaler::Formatter = scaler::Formatter::new()
   .set_scaling(scaler::Scaling::None)
   .set_separators(".", ",");
assert_eq!(f.format(1), "1,000");
assert_eq!(f.format(10), "10,00");
assert_eq!(f.format(100), "100,0");
assert_eq!(f.format(1000), "1.000");
assert_eq!(f.format(10000), "10.000");

let f: scaler::Formatter = scaler::Formatter::new()
   .set_scaling(scaler::Scaling::None)
   .set_separators("", ",");
assert_eq!(f.format(1), "1,000");
assert_eq!(f.format(10), "10,00");
assert_eq!(f.format(100), "100,0");
assert_eq!(f.format(1000), "1000");
assert_eq!(f.format(10000), "10000");

let f: scaler::Formatter = scaler::Formatter::new()
   .set_scaling(scaler::Scaling::None)
   .set_separators(",", ".");
assert_eq!(f.format(1), "1.000");
assert_eq!(f.format(10), "10.00");
assert_eq!(f.format(100), "100.0");
assert_eq!(f.format(1000), "1,000");
assert_eq!(f.format(10000), "10,000");

Sign

• Always

  • Always show sign, even when number is positive.

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_sign(scaler::Sign::Always);
  assert_eq!(f.format(std::f64::NEG_INFINITY), "-∞");
  assert_eq!(f.format(-1), "-1,000");
  assert_eq!(f.format(0), "+0,000");
  assert_eq!(f.format(1), "+1,000");
  assert_eq!(f.format(std::f64::INFINITY), "+∞");

• OnlyMinus

  • Only show sign when number is negative.

  let f: scaler::Formatter = scaler::Formatter::new()
     .set_sign(scaler::Sign::OnlyMinus);
  assert_eq!(f.format(std::f64::NEG_INFINITY), "-∞");
  assert_eq!(f.format(-1), "-1,000");
  assert_eq!(f.format(0), "0,000");
  assert_eq!(f.format(1), "1,000");
  assert_eq!(f.format(std::f64::INFINITY), "∞");

scaler's Issues

Option to discard trailing Zeros

Add an Option to the Formatter to discard trailing zeros.
May be with an option to keep n zeros.