/expression-calculator

A calculator that calculates expressions like add(1, mult(2, 3)), or let(a, 5, add(a, a))

Primary LanguageJavaDo What The F*ck You Want To Public LicenseWTFPL

expression-calculator

Java requirement

A calculator that calculates expressions like add(1, mult(2, 3)), or let(a, 5, add(a, a)).

Build

Install Maven and execute mvn clean package -P default,withTests.

Usage

./exprcalculator.sh [options] "expression"
  Options:
    -v, -log
      Logging level. Possible values: OFF, ERROR, WARN, INFO, DEBUG
      Default: ERROR
    -p, -precision
      Precision: the number of digits to be used. Must not be negative. Use 0 for unlimited
      precision, but this will result in failures to calculate expressions which lead to irrational
      or repeating decimals (e.g. 'div(1, 3)')
      Default: precision=7 roundingMode=HALF_EVEN

One can also specify expression via the standard input stream:

echo "expression" | ./exprcalculator.sh

If you experience a slow startup (about a second) just know that this is caused by the initialization done by the logging library. An insanity that I would have solved if it were a production-ready command-line tool.

Expression systax

Syntax is pretty much obvious from the examples below, but some notes still might be helphul:

  • An expression is either <number>, or <any operator>(<operand1>, <operand2>, ...), or <declared variable name>.
  • Birary operators: add, sub, mult, div: <operator>(<expression>, <expression>).
  • Ternary operator let allows declaring variables and assign values for them: let(<variable name>, <variable value expression>, <expression where variable is used>). Names of variables can not be the same as names of operators.

Examples

./exprcalculator.sh "-3.14"
-3.14

The next calculation fails because (-3.14) is not an expression:

./exprcalculator.sh "(-3.14)"
Problem with ')' at index 6:
(-3.14)
      ^

./exprcalculator.sh "add(1, 2)"
3

Unlimited precision is specified with -p 0, but 0.1 divided by 0.3 is a repeating decimal and can not be represented with unlimited precision:

./exprcalculator.sh -p 0 "div(0.1, 0.3)"
Problem with 'div' at index 0:
div(0.1, 0.3)
^

./exprcalculator.sh "add(1, mult(2, 3))"
7

./exprcalculator.sh "mult(add(2, 2), div(9, 3))"
12

./exprcalculator.sh "let(a, 5, add(a, a))"
10

./exprcalculator.sh "let(a, 5, let(b, mult(a, 10), add(b, a)))"
55

./exprcalculator.sh "let(a, let(b, 10, add(b, b)), let(b, 20, add(a, b)))"
40

./exprcalculator.sh "let(a, let(b, let(_var, 1, div(1, _var)), add(b, b)), let(b, 20, add(a, b)))"
22

./exprcalculator.sh "let(a, 1, add(let(b, 2, div(b, 1)), -1.0))"
1.0

This calculation fails because we try to define var twice within the same scope:

./exprcalculator.sh "let(var, 1, add(let(var, 2, div(var, 1)), 1))"
Problem with 'var' at index 20:
let(var, 1, add(let(var, 2, div(var, 1)), 1))
                    ^

Licensed under WTFPL, except where another license is explicitly specified.