Handling errors in the Go API | CUE

Paul Jolly

Jonathan Matthews

go api

The CUE Go API contains several functions that might need to communicate runtime errors to their caller, such as problems during evaluation or validation. They do this using the cue/errors.Error type, and the cue/errors package contains functions that allow you to interrogate and manipulate these errors.

This guide demonstrates how to catch these errors and how to use the cue/errors package to examine their details.

Set up some CUE code

If you don’t already have an existing CUE module, initialize one to contain a configuration:

TERMINAL

$ cue mod init cue.example

If you don’t already have some known-bad CUE that contains evaluation or validation errors, create some:

package example

#Schema: {
	i: int
	s: string
}

val: #Schema & {
	i: "hello" // Not an int.
	s: 42      // Not a string.
}

Verify that the configuration does not validate successfully, as expected:

TERMINAL

$ cue vet
val.i: conflicting values int and "hello" (mismatched types int and string):
    ./bad.cue:4:5
    ./bad.cue:9:5
val.s: conflicting values string and 42 (mismatched types string and int):
    ./bad.cue:5:5
    ./bad.cue:10:5

Set up a Go program

If you don’t already have an existing Go module, initialize one to contain a Go program:

TERMINAL

$ go mod init go.example
...

Write a Go program that loads the CUE, evalutes some of it, and then prints a summary and details of any evaluation or validation errors that ocurr:

package main

import (
	"fmt"

	"cuelang.org/go/cue"
	"cuelang.org/go/cue/cuecontext"
	"cuelang.org/go/cue/errors"
	"cuelang.org/go/cue/load"
)

func main() {
	ctx := cuecontext.New()

	// Load the CUE package in the current directory and extract a value from it.
	insts := load.Instances([]string{"."}, nil)
	pkg := ctx.BuildInstance(insts[0])
	val := pkg.LookupPath(cue.ParsePath("val"))

	// Check if the value evaluates successfully.
	err := val.Validate(cue.Concrete(true))
	if err != nil {
		// Extract one or more errors from the evaluation.
		errs := errors.Errors(err)

		// Display information about the errors.
		fmt.Println(`# Error summary [err]:`)
		fmt.Printf("%v\n\n", err)

		fmt.Println(`# Error details [errors.Details(err)]:`)
		fmt.Printf("%v\n", errors.Details(err, nil))

		fmt.Println(`# Error count [len(errs)]:`)
		fmt.Printf("%d\n", len(errs))
	}
}

Add a dependency on cuelang.org/go and ensure the Go module is tidy:

TERMINAL

$ go get cuelang.org/go@v0.11.1
...
$ go mod tidy
...

You can use @latest in place of the specific version mentioned here.

Run the Go program

Run the Go program:

TERMINAL

$ go run .
# Error summary [err]:
val.i: conflicting values int and "hello" (mismatched types int and string) (and 1 more errors)

# Error details [errors.Details(err)]:
val.i: conflicting values int and "hello" (mismatched types int and string):
    /home/runner/bad.cue:4:5
    /home/runner/bad.cue:9:5
val.s: conflicting values string and 42 (mismatched types string and int):
    /home/runner/bad.cue:5:5
    /home/runner/bad.cue:10:5

# Error count [len(errs)]:
2

Notice that cue/errors.Details is simply a convenience function that returns the error texts of the underlying errors, aggregated as a single string. The information and metadata contained in each underlying error can be accessed by iterating through the slice of individual errors returned by the cue/errors.Errors method.

Tag: go api – pages documenting and using the Go API
Go API: cue/errors – package documentation

Set up some CUE code

Set up a Go program

Run the Go program

Related content