DavudSafarli / design-calculator-challenge

Calculator

Simple calculator application, written in Go. Supports basic BODMAS operations.

Example usage:

func main() {
	c := calculator.New()
	res, err := c.Eval("5+5")
	fmt.Println(res, err) // 10, <nil>
}

Handling error cases:

func main() {
	c := calculator.New()
	res, err := c.Eval("?5+5")
	fmt.Println(res, err) // 0, unknown character '?'

	if e, ok := err.(lexer.UnknownSymbolError); ok {
		fmt.Printf("cannot use %q symbol\n", e.Symbol)
	}
}

func main() {
	c := calculator.New()

	input := "5+*5"
	res, err := c.Eval(input)
	fmt.Println(res, err) // 0, cannot have 2 operators side by side in position (2, 3)

	if evalErr, ok := err.(calculator.EvalError); ok {
        // show invalid char with 1 preceding and succeeding characters
		from := max(0, evalErr.StartPos-1)
		to := min(len(input), evalErr.EndPos+1)
		invalidPart := input[from:to]
		fmt.Printf("%v; %q", evalErr.Err, invalidPart) // cannot have 2 operators side by side; "+*5"
	}
}

func max(a, b int) int {
	if a > b {
		return a
	}
	return b
}
func min(a, b int) int {
	if a < b {
		return a
	}
	return b
}

How it works

It goes through multiple steps to calculate the input expression.

1. First step is lexical analysis of the given input, and converting it into a slice of Token. You can think of a Token as a keyword, or a character that has a meaning. It can also hold any value related to itself. In the context of this application, some of the Tokens are :

   • NUM
   • ADD
   • SUB
   • MUL
   • DIV
   • POW
   • L_PAR
   • R_PAR

   As an example, expression "1+2" after lexical analysis produces: [Token(NUM, 1), Token(ADD), Token(NUM, 2)]

2. Second step is validating the tokenized input find and report errors and error positions. For example: "(*4" -> [LEFT_PAR, MUL, NUM] is an invalid expression. The position(index) where it starts being invalid is 1(MUL), and ends at 2, because we cannot have Multiplication after the opening-paracantesis. Here, we report the Error(ErrOperationAfterLeftParacantesis), the start and endposition, so that we can show the incorrect part to the end-user.

3. Parsing the Tokens, and building the Expression Tree. The way we build the expression tree is parsing the Tokens from Infix notation to Postfix and then, Postfix to Expression Tree. Here in this application, this steps happen at the same time. You can learn more about Infix to Postfix from this video. You can read this geeksforgeeks article on Program to convert Infix notation to Expression Tree to get learn more about the algorithm.

4. Evaluating the Expression tree recursively and getting the result.