checkpoint

Documentation

Overview

Package checkpoint provides a way to decorate errors by some additional caller information which results in something similar to a stacktrace. Each error added to a checkpoint can be checked by errors.Is and retrieved by errors.As.

Index

• func From(err error) error
• func Wrap(prev, err error) error

Functions

func From

func From(err error) error

From just wraps an error by a new checkpoint which adds some caller information to the error. It returns nil, if err == nil.

func Wrap

func Wrap(prev, err error) error

Wrap adds a checkpoint with some caller information from an error and accepts also another error which can further describe the checkpoint. Returns nil if prev == nil. If err is nil, it still creates a checkpoint. This allows for example to predefine some errors and use them later:

var(
		ErrSomethingSpecialWentWrong = errors.New("a very bad error")
)
func someFunction() error {
	err := somethingOtherThatThrowsErrors()
	return checkpoint.Wrap(err, ErrSomethingSpecialWentWrong)
}

err := someFunction()

If used that way, you can still check with errors.Is() for the ErrSomethingSpecialWentWrong

if errors.Is(err, ErrSomethingSpecialWentWrong) {
	fmt.Println("The special error was thrown")
} else {
	fmt.Println(err)
}

but also for the error returned by somethingOtherThatThrowsErrors() (if you know what error it is). If the error in this example is nil, no checkpoint gets created.