Search⌘ K
AI Features
Home
Courses
Learning Test-Driven Development with Go

Error Handling

Explore how Go handles errors as values and learn methods to define errors, including implementing the error interface and using errors.New and fmt.Errorf. Understand control flow with defer, panic, and recover keywords and how to wrap and chain errors for better context in your programs.

We'll cover the following...

Errors in Go

Errors are normal values within Go. We can deal with them in the same way we deal with other types, such as int, string, bool, and so on.

How we can define an error

There are three different ways we can define an error:

  • By implementing the error interface
  • By calling the errors.New() function
  • By calling the fmt.Errorf() function

Let’s see the first two methods in action in the code snippet below:

Go (1.18.2)
package main
import (
	"errors"
	"fmt"
)
// 1. defining struct for error
type CustomErr struct {
	Code     string
	ErrorMsg string
}
// 2. implement error interface
func (c *CustomErr) Error() string {
	return fmt.Sprintf("code: %q\nmessage: %q", c.Code, c.ErrorMsg)
}
func Divide(a, b int) (int, error) {
	if b == 0 {
		// 3. return our custom error
		return 0, &CustomErr{Code: "INVALID", ErrorMsg: "cannot divide by 0"}
	}
	return a / b, nil
}
func main() {
	a, b := 3, 0
	_, err := Divide(a, b)
	if err != nil {
		// 4. run the Error() method on error returned
		fmt.Println(err.Error())
		// 5. other method to create an error
		err = errors.New("sample error")
		fmt.Println(err.Error())
		return
	}
}

In the example, ...