Clock as struct

// Package Clock is a small library for keeping track of time.
package clock

import (
	"fmt"
)

const (
	dayMinutes  = 1440
	hourMinutes = 60
)

// Clock type
type Clock struct {
	minutes int
}

func normalize(minutes int) Clock {
	if minutes < 0 {
		return Clock{dayMinutes + (minutes % dayMinutes)}
	} else if minutes >= dayMinutes {
		return Clock{minutes % dayMinutes}
	}
	return Clock{minutes}
}

// New returns a Clock that represents the hour and minutes.
func New(hour, minutes int) Clock {
	return normalize(hour*hourMinutes + minutes)
}

// Add returns a Clock that represents the Clock plus minutes.
func (c Clock) Add(minutes int) Clock {
	return normalize(c.minutes + minutes)
}

// Subtract returns a Clock that represents the Clock minus minutes.
func (c Clock) Subtract(minutes int) Clock {
	return normalize(c.minutes - minutes)
}

// String returns the Clock formatted into hours:minutes, e.g. 01:02.
func (c *Clock) String() string {
	return fmt.Sprintf("%02d:%02d", c.minutes/hourMinutes, c.minutes%hourMinutes)
}

The approach imports the fmt package so it can use the fmt.Sprintf() function in the String() method.

This approach starts by typing Clock as a struct with an int field.

It then defines a couple of constants with meaningful names for 1440 and 60, instead of using the literals as magic numbers.

The normalize() helper function is defined to properly handle rollover for when adding or subtracting minutes crosses into the next or previous day(s). There are other ways to handle rollover, such as the following expression:

((hours*hourMinutes+minutes)%dayminutes + dayMinutes) % dayMinutes

Since normalize() returns a Clock, it can be used by the New function as well as the Add and Subtract methods.

The String() function uses fmt.Sprintf() to format the hour and minutes with leading zeros. If the amount of minutes is 69, then the hour is 69 divided by 60 (1), and the minutes is the remainder of 69 divided by 60 (9), so the string for 69 minutes would be 01:09. The format specifier %02d says to pad a base-10 number with zeros for up to two places. If one place is occupied by a digit, then the extra place is padded with a zero. More info can be found at a format cheat sheet.