Lib.rs

RionaCalc

Mathematical expression calculator with big integers, floats, common fractions, and complex numbers support. It can be launched in two modes: immediate, when rclc gets an expression, calculates it, prints the result, and exits; and interactive, when rclc displays a prompt and calculates every expression you type.

Suggestions, ideas, pull requests are very welcome. Thank you!

Why yet another calculator

From time to time I need to evaluate an expression that contains values other than float point numbers: common fractions, complex numbers, arbitrary precision integer numbers. But majority of calculators works only with floating point numbers. Even if a calculator supports any of other types of numbers, it often requires switching to special modes. E.g, calculators that supported common fractions required to enter common-fraction mode to use them - in this mode other kinds of numbers are unavailable. Some calculators support complex numbers but do not do it transparently for a user. E.g, its documentation states that complex numbers are supported but a calculator fails on a evaluating square root of negative number with an error "Invalid argument". I wanted to have a calculator that allows me to mix any types of numbers in one expression and it is able to detect the correct type of the result and argument.

Lesser requirements:

• evaluate trigonometric functions using radians and degrees without turning radians/degrees switch on and off
• a calculator that works in terminal either in interactive mode or is able to calculate an expression passed in command line and display the result. Yes, bc can do it. But its usage seems counterintuitive to me: instead of simple call bc "expr" it must be launched as echo "expr" | bc
• user-defined variables. Though, majority of calculators supports it out of the box

Very simple examples(ans - an special variable that holds the result of the last successful evaluation):

> sqrt(-2)  // square root of negative number
= 0.0+1.4142135623730952i
> sqr(ans) // square root of a complex number may produce real number
= -2.0000000000000006
> 345**12 // big integer in action
= 2843342266303054544082275390625
> 1\2 + 3\5  // one half and three fifth is one and one tenth
= 1\1\10
> sqr(3\5)  // square of a rational number is a rational number
= 9\25
> sin(90°) == sin(pi/2) // degrees and radians mixed in one expression, '°' can be replaced with 'd' for easier typing 
= 1

Features

• No modes: all types of values can be transparently used in one expression. E.g, (1\2 + 3\5) * 2-3i + sin(30d) + cos(0.25) - multiply sum of two common fractions - one half and three fifth - by a complex number, add sine of 30 degrees and cosine of 0.25 radians. Spaces are added only for readability, they can be omitted
• Automatic selection of more appropriate argument type for a function: e.g, sqrt(-4) converts float number -4 into complex one -4+0i and then calculates the result 0+2i. The same is true for calculating logarithm for negative float numbers, and acos and asin for argument greater than 1.0
• Automatic adding multiplication sign where it is omitted: e.g, (1+2)(2+9) is calculated as (1+2)*(2+9)
• Functions with a single-value argument do not require to enclose its argument into brackets: e.g, sin cos 2 is calculated as sin(cos(2))
• The final closing brackets can be omitted: e.g, (1+2)*(2+9 is the same as (1+2)*(2+9)
• Trigonometric functions work with radians and degrees. Bare numbers are treated as radians, degrees requires one or three suffixes. Two degrees formats: 20d30m50s or 20°30'50". Minutes and seconds can be omitted, in this case degrees can be float number like 30.25d. So, sin(pi/2) == sin(90°)
• Every number can include group separator _ for readability - it is very useful when using big integers. 3_000.90_23 == 3000.9023
• Both . and , are treated as decimal separators
• Function argument separator is ;. If a function receives more arguments than it requires, the trailing arguments are dropped: e.g, sqrt(11;12;13) is the same as sqrt(11)
• Regular fractions use \ to separate its parts. They can be written with integer part or only with numerator and denominator, e.g 1\1\10 == 11\10
• Two complex numbers formats: with marker at the end or in the middle. E.g, 1+2i == 1+i2. In addition, j can be used instead of i - but the calculator outputs always with i
• Hexadecimal(starts with 0x), octal(starts with 0o), and binary(starts with 0b) numbers
• Basic variable and scripting support allows users to create their own constant libraries and preload them at calculator startup
• Commands in interactive mode(a very limited set at this moment): quit or exit close the calculator, and load <filename> - load the file and evaluate lines one by one, skipping comments, the last evaluated result is printed
• Character % can be either a modulo or a percentage operator. It depends on the character position: if % is right before the expression end or before closing bracket or before another operator and previous operator is one of +, -, *, or /, the character is considered a percentage operator

Please, read the detailed documentation here.

Installation

The application can be compiled from source, or installed using cargo:

$ cargo install rclc

You need Rust compiler that supports Rust 2018 edition (Rust 1.31 or newer) to do it. If you want to upgrade existing rclc, execute the following command:

$ cargo install rclc --force

Pre-compiled binaries

For Windows you can download pre-compiled binaries from Release page.

• Windows binary works on Windows 7 or newer Windows.

Known issues

• The calculator is not thoroughly tested, bugs may happen. Please, notify me about any issue
• Float numbers with arbitrary precision are not supported yet - all floats are 64-bit float numbers. I am aware of rust-port of GNU GMP, but I do not want to use it at this moment - I remember having troubles trying to build the library on Windows